Skip to contentSkip to conversation

media: davinci-isif: avoid uninitialized variable use

PATCH In review Sasha Levin 5 participants 416 messages Jan 16, 2020 – Jan 23, 2020 Page 2 of 3

Subsystems (59)
lore
Share & tools
lore
author maintainer
#172 lore raw

PATCH 436/671 nvmem: imx-ocotp: Change TIMING calculation to u-boot algorithm

In reply to Sasha Levin

From: Bryan O'Donoghue <[email protected]>

[ Upstream commit 159dbaf57b2f4f67ecb59b2c87d071e45ed41d7e ]

The RELAX field of the OCOTP block is turning out as a zero on i.MX8MM. This messes up the subsequent re-load of the fuse shadow registers.

After some discussion with people @ NXP its clear we have missed a trick here in Linux.

The OCOTP fuse programming time has a physical minimum 'burn time' that is not related to the ipg_clk.

We need to define the RELAX, STROBE_READ and STROBE_PROG fields in terms of desired timings to allow for the burn-in to safely complete. Right now only the RELAX field is calculated in terms of an absolute time and we are ending up with a value of zero.

This patch inherits the u-boot timings for the OCOTP_TIMING calculation on the i.MX6 and i.MX8. Those timings are known to work and critically specify values such as STROBE_PROG as a minimum timing.

Fixes 0642bac7da42 ("nvmem: imx-ocotp: add write support")
---
 drivers/nvmem/imx-ocotp.c | 36 ++++++++++++++++++++++++++++++++----
 1 file changed, 32 insertions(+), 4 deletions(-)
drivers/nvmem/imx-ocotp.c +32 -4
index 04421a73f74a..09281aca86c2 100644
@@ -50,7 +50,9 @@
 #define IMX_OCOTP_BM_CTRL_ERROR		0x00000200
 #define IMX_OCOTP_BM_CTRL_REL_SHADOWS	0x00000400
 
-#define DEF_RELAX			20	/* > 16.5ns */
+#define TIMING_STROBE_PROG_US		10	/* Min time to blow a fuse */
+#define TIMING_STROBE_READ_NS		37	/* Min time before read */
+#define TIMING_RELAX_NS			17
 #define DEF_FSOURCE			1001	/* > 1000 ns */
 #define DEF_STROBE_PROG			10000	/* IPG clocks */
 #define IMX_OCOTP_WR_UNLOCK		0x3E770000
@@ -182,12 +184,38 @@ static void imx_ocotp_set_imx6_timing(struct ocotp_priv *priv)
 	 * fields with timing values to match the current frequency of the
 	 * ipg_clk. OTP writes will work at maximum bus frequencies as long
 	 * as the HW_OCOTP_TIMING parameters are set correctly.
+	 *
+	 * Note: there are minimum timings required to ensure an OTP fuse burns
+	 * correctly that are independent of the ipg_clk. Those values are not
+	 * formally documented anywhere however, working from the minimum
+	 * timings given in u-boot we can say:
+	 *
+	 * - Minimum STROBE_PROG time is 10 microseconds. Intuitively 10
+	 *   microseconds feels about right as representative of a minimum time
+	 *   to physically burn out a fuse.
+	 *
+	 * - Minimum STROBE_READ i.e. the time to wait post OTP fuse burn before
+	 *   performing another read is 37 nanoseconds
+	 *
+	 * - Minimum RELAX timing is 17 nanoseconds. This final RELAX minimum
+	 *   timing is not entirely clear the documentation says "This
+	 *   count value specifies the time to add to all default timing
+	 *   parameters other than the Tpgm and Trd. It is given in number
+	 *   of ipg_clk periods." where Tpgm and Trd refer to STROBE_PROG
+	 *   and STROBE_READ respectively. What the other timing parameters
+	 *   are though, is not specified. Experience shows a zero RELAX
+	 *   value will mess up a re-load of the shadow registers post OTP
+	 *   burn.
 	 */
 	clk_rate = clk_get_rate(priv->clk);
 
-	relax = clk_rate / (1000000000 / DEF_RELAX) - 1;
-	strobe_prog = clk_rate / (1000000000 / 10000) + 2 * (DEF_RELAX + 1) - 1;
-	strobe_read = clk_rate / (1000000000 / 40) + 2 * (DEF_RELAX + 1) - 1;
+	relax = DIV_ROUND_UP(clk_rate * TIMING_RELAX_NS, 1000000000) - 1;
+	strobe_read = DIV_ROUND_UP(clk_rate * TIMING_STROBE_READ_NS,
+				   1000000000);
+	strobe_read += 2 * (relax + 1) - 1;
+	strobe_prog = DIV_ROUND_CLOSEST(clk_rate * TIMING_STROBE_PROG_US,
+					1000000);
+	strobe_prog += 2 * (relax + 1) - 1;
 
 	timing = readl(priv->base + IMX_OCOTP_ADDR_TIMING) & 0x0FC00000;
 	timing |= strobe_prog & 0x00000FFF;
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#173 lore raw

PATCH 437/671 tools: bpftool: use correct argument in cgroup errors

In reply to Sasha Levin

From: Jakub Kicinski <[email protected]>

[ Upstream commit 6c6874f401e5a0caab3b6a0663169e1fb5e930bb ]

cgroup code tries to use argv[0] as the cgroup path, but if it fails uses argv[1] to report errors.

Fixes 5ccda64d38cc ("bpftool: implement cgroup bpf operations")
---
 tools/bpf/bpftool/cgroup.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
tools/bpf/bpftool/cgroup.c +3 -3
index ee7a9765c6b3..adbcd84818f7 100644
@@ -164,7 +164,7 @@ static int do_show(int argc, char **argv)
 
 	cgroup_fd = open(argv[0], O_RDONLY);
 	if (cgroup_fd < 0) {
-		p_err("can't open cgroup %s", argv[1]);
+		p_err("can't open cgroup %s", argv[0]);
 		goto exit;
 	}
 
@@ -345,7 +345,7 @@ static int do_attach(int argc, char **argv)
 
 	cgroup_fd = open(argv[0], O_RDONLY);
 	if (cgroup_fd < 0) {
-		p_err("can't open cgroup %s", argv[1]);
+		p_err("can't open cgroup %s", argv[0]);
 		goto exit;
 	}
 
@@ -403,7 +403,7 @@ static int do_detach(int argc, char **argv)
 
 	cgroup_fd = open(argv[0], O_RDONLY);
 	if (cgroup_fd < 0) {
-		p_err("can't open cgroup %s", argv[1]);
+		p_err("can't open cgroup %s", argv[0]);
 		goto exit;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#174 lore raw

PATCH 438/671 backlight: pwm_bl: Fix heuristic to determine number of brightness levels

In reply to Sasha Levin

From: Matthias Kaehlcke <[email protected]>

[ Upstream commit 73fbfc499448455f1e1c77717040e09e25f1d976 ]

With commit 88ba95bedb79 ("backlight: pwm_bl: Compute brightness of LED linearly to human eye") the number of set bits (aka hweight()) in the PWM period is used in the heuristic to determine the number of brightness levels, when the brightness table isn't specified in the DT. The number of set bits doesn't provide a reliable clue about the length of the period, instead change the heuristic to:

nlevels = period / fls(period)

Also limit the maximum number of brightness levels to 4096 to avoid excessively large tables.

With this the number of levels increases monotonically with the PWM period, until the maximum of 4096 levels is reached:

period (ns) # levels

100 16
500 62
1000 111
5000 416
10000 769
50000 3333
100000 4096

Fixes 88ba95bedb79 ("backlight: pwm_bl: Compute brightness of LED linearly to human eye")
---
 drivers/video/backlight/pwm_bl.c | 24 ++++++------------------
 1 file changed, 6 insertions(+), 18 deletions(-)
drivers/video/backlight/pwm_bl.c +5 -18
index 7ddc0930e98c..3a3098d4873b 100644
@@ -199,29 +199,17 @@ int pwm_backlight_brightness_default(struct device *dev,
 				     struct platform_pwm_backlight_data *data,
 				     unsigned int period)
 {
-	unsigned int counter = 0;
-	unsigned int i, n;
+	unsigned int i;
 	u64 retval;
 
 	/*
-	 * Count the number of bits needed to represent the period number. The
-	 * number of bits is used to calculate the number of levels used for the
-	 * brightness-levels table, the purpose of this calculation is have a
-	 * pre-computed table with enough levels to get linear brightness
-	 * perception. The period is divided by the number of bits so for a
-	 * 8-bit PWM we have 255 / 8 = 32 brightness levels or for a 16-bit PWM
-	 * we have 65535 / 16 = 4096 brightness levels.
-	 *
-	 * Note that this method is based on empirical testing on different
-	 * devices with PWM of 8 and 16 bits of resolution.
+	 * Once we have 4096 levels there's little point going much higher...
+	 * neither interactive sliders nor animation benefits from having
+	 * more values in the table.
 	 */
-	n = period;
-	while (n) {
-		counter += n % 2;
-		n >>= 1;
-	}
+	data->max_brightness +		min((int)DIV_ROUND_UP(period, fls(period)), 4096);
 
-	data->max_brightness = DIV_ROUND_UP(period, counter);
 	data->levels = devm_kcalloc(dev, data->max_brightness,
 				    sizeof(*data->levels), GFP_KERNEL);
 	if (!data->levels)
Signature · 1 line
2.20.1
author maintainer
#175 lore raw

PATCH 439/671 fork,memcg: alloc_thread_stack_node needs to set tsk->stack

In reply to Sasha Levin

From: Andrea Arcangeli <[email protected]>

[ Upstream commit 1bf4580e00a248a2c86269125390eb3648e1877c ]

Commit 5eed6f1dff87 ("fork,memcg: fix crash in free_thread_stack on memcg charge fail") corrected two instances, but there was a third instance of this bug.

Without setting tsk->stack, if memcg_charge_kernel_stack fails, it'll execute free_thread_stack() on a dangling pointer.

Enterprise kernels are compiled with VMAP_STACK=y so this isn't critical, but custom VMAP_STACK=n builds should have some performance advantage, with the drawback of risking to fail fork because compaction didn't succeed. So as long as VMAP_STACK=n is a supported option it's worth fixing it upstream.

---
 kernel/fork.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
kernel/fork.c +5 -1
index 1bd119530a49..1a2d18e98bf9 100644
@@ -240,7 +240,11 @@ static unsigned long *alloc_thread_stack_node(struct task_struct *tsk, int node)
 	struct page *page = alloc_pages_node(node, THREADINFO_GFP,
 					     THREAD_SIZE_ORDER);
 
-	return page ? page_address(page) : NULL;
+	if (likely(page)) {
+		tsk->stack = page_address(page);
+		return tsk->stack;
+	}
+	return NULL;
 #endif
 }
 
Signature · 1 line
2.20.1
author maintainer
#176 lore raw

PATCH 440/671 bnxt_en: Fix ethtool selftest crash under error conditions.

In reply to Sasha Levin

From: Michael Chan <[email protected]>

[ Upstream commit d27e2ca1166aefd54d9c48fb6647dee8115a5dfc ]

After ethtool loopback packet tests, we re-open the nic for the next IRQ test. If the open fails, we must not proceed with the IRQ test or we will crash with NULL pointer dereference. Fix it by checking the bnxt_open_nic() return code before proceeding.

Fixes 67fea463fd87 ("bnxt_en: Add interrupt test to ethtool -t selftest.")
---
 drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +3 -3
index 0a409ba4012a..dc63d269f01d 100644
@@ -2600,7 +2600,7 @@ static void bnxt_self_test(struct net_device *dev, struct ethtool_test *etest,
 	bool offline = false;
 	u8 test_results = 0;
 	u8 test_mask = 0;
-	int rc, i;
+	int rc = 0, i;
 
 	if (!bp->num_tests || !BNXT_SINGLE_PF(bp))
 		return;
@@ -2671,9 +2671,9 @@ static void bnxt_self_test(struct net_device *dev, struct ethtool_test *etest,
 		}
 		bnxt_hwrm_phy_loopback(bp, false, false);
 		bnxt_half_close_nic(bp);
-		bnxt_open_nic(bp, false, true);
+		rc = bnxt_open_nic(bp, false, true);
 	}
-	if (bnxt_test_irq(bp)) {
+	if (rc || bnxt_test_irq(bp)) {
 		buf[BNXT_IRQ_TEST_IDX] = 1;
 		etest->flags |= ETH_TEST_FL_FAILED;
 	}
Signature · 1 line
2.20.1
author maintainer
#177 lore raw

PATCH 441/671 bnxt_en: Suppress error messages when querying DSCP DCB capabilities.

In reply to Sasha Levin

From: Michael Chan <[email protected]>

[ Upstream commit 4ca5fa39e1aea2f85eb9c4257075c4077c6531da ]

Some firmware versions do not support this so use the silent variant to send the message to firmware to suppress the harmless error. This error message is unnecessarily alarming the user.

Fixes afdc8a84844a ("bnxt_en: Add DCBNL DSCP application protocol support.")
---
 drivers/net/ethernet/broadcom/bnxt/bnxt_dcb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_dcb.c +1 -1
index a85d2be986af..0e4e0b47f5d8 100644
@@ -396,7 +396,7 @@ static int bnxt_hwrm_queue_dscp_qcaps(struct bnxt *bp)
 
 	bnxt_hwrm_cmd_hdr_init(bp, &req, HWRM_QUEUE_DSCP_QCAPS, -1, -1);
 	mutex_lock(&bp->hwrm_cmd_lock);
-	rc = _hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
+	rc = _hwrm_send_message_silent(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
 	if (!rc) {
 		bp->max_dscp_value = (1 << resp->num_dscp_bits) - 1;
 		if (bp->max_dscp_value < 0x3f)
Signature · 1 line
2.20.1
author maintainer
#178 lore raw

PATCH 442/671 iommu/amd: Make iommu_disable safer

In reply to Sasha Levin

From: Kevin Mitchell <[email protected]>

[ Upstream commit 3ddbe913e55516d3e2165d43d4d5570761769878 ]

Make it safe to call iommu_disable during early init error conditions before mmio_base is set, but after the struct amd_iommu has been added to the amd_iommu_list. For example, this happens if firmware fails to fill in mmio_phys in the ACPI table leading to a NULL pointer dereference in iommu_feature_disable.

Fixes 2c0ae1720c09c ('iommu/amd: Convert iommu initialization to state machine')
---
 drivers/iommu/amd_iommu_init.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/iommu/amd_iommu_init.c +3 -0
index 1e9a5da562f0..465f28a7844c 100644
@@ -422,6 +422,9 @@ static void iommu_enable(struct amd_iommu *iommu)
 
 static void iommu_disable(struct amd_iommu *iommu)
 {
+	if (!iommu->mmio_base)
+		return;
+
 	/* Disable command buffer */
 	iommu_feature_disable(iommu, CONTROL_CMDBUF_EN);
 
Signature · 6 lines
2.20.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author maintainer
#179 lore raw

PATCH 443/671 mfd: intel-lpss: Release IDA resources

In reply to Sasha Levin

From: Andy Shevchenko <[email protected]>

[ Upstream commit 02f36911c1b41fcd8779fa0c135aab0554333fa5 ]

ida instances allocate some internal memory for ->free_bitmap in addition to the base 'struct ida'. Use ida_destroy() to release that memory at module_exit().

Fixes 4b45efe85263 ("mfd: Add support for Intel Sunrisepoint LPSS devices")
---
 drivers/mfd/intel-lpss.c | 1 +
 1 file changed, 1 insertion(+)
drivers/mfd/intel-lpss.c +1 -0
index ff3fba16e735..95e217e6b6d7 100644
@@ -545,6 +545,7 @@ module_init(intel_lpss_init);
 
 static void __exit intel_lpss_exit(void)
 {
+	ida_destroy(&intel_lpss_devid_ida);
 	debugfs_remove(intel_lpss_debugfs);
 }
 module_exit(intel_lpss_exit);
Signature · 1 line
2.20.1
author maintainer
#180 lore raw

PATCH 444/671 rxrpc: Fix uninitialized error code in rxrpc_send_data_packet()

In reply to Sasha Levin

From: David Howells <[email protected]>

[ Upstream commit 3427beb6375d04e9627c67343872e79341a684ea ]

With gcc 4.1:

net/rxrpc/output.c: In function ‘rxrpc_send_data_packet’:
net/rxrpc/output.c:338: warning: ‘ret’ may be used uninitialized in this function

Indeed, if the first jump to the send_fragmentable label is made, and the address family is not handled in the switch() statement, ret will be used uninitialized.

Fix this by BUG()'ing as is done in other places in rxrpc where internal support for future address families will need adding. It should not be possible to reach this normally as the address families are checked up-front.

Fixes 5a924b8951f835b5 ("rxrpc: Don't store the rxrpc header in the Tx queue sk_buffs")
---
 net/rxrpc/output.c | 3 +++
 1 file changed, 3 insertions(+)
net/rxrpc/output.c +3 -0
index 345dc1c5fe72..31e47cfb3e68 100644
@@ -524,6 +524,9 @@ int rxrpc_send_data_packet(struct rxrpc_call *call, struct sk_buff *skb,
 		}
 		break;
 #endif
+
+	default:
+		BUG();
 	}
 
 	if (ret < 0)
Signature · 1 line
2.20.1
author maintainer
#181 lore raw

PATCH 445/671 xprtrdma: Fix use-after-free in rpcrdma_post_recvs

In reply to Sasha Levin

From: Chuck Lever <[email protected]>

[ Upstream commit 2d0abe36cf13fb7b577949fd1539326adddcc9bc ]

Dereference wr->next /before/ the memory backing wr has been released. This issue was found by code inspection. It is not expected to be a significant problem because it is in an error path that is almost never executed.

Fixes 7c8d9e7c8863 ("xprtrdma: Move Receive posting to ... ")
---
 net/sunrpc/xprtrdma/verbs.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
net/sunrpc/xprtrdma/verbs.c +2 -1
index 5ddbf227e7c6..2a1d8ec7f706 100644
@@ -1558,10 +1558,11 @@ rpcrdma_post_recvs(struct rpcrdma_xprt *r_xprt, bool temp)
 	rc = ib_post_recv(r_xprt->rx_ia.ri_id->qp, wr,
 			  (const struct ib_recv_wr **)&bad_wr);
 	if (rc) {
-		for (wr = bad_wr; wr; wr = wr->next) {
+		for (wr = bad_wr; wr;) {
 			struct rpcrdma_rep *rep;
 
 			rep = container_of(wr, struct rpcrdma_rep, rr_recv_wr);
+			wr = wr->next;
 			rpcrdma_recv_buffer_put(rep);
 			--count;
 		}
Signature · 1 line
2.20.1
author maintainer
#182 lore raw

PATCH 446/671 um: Fix IRQ controller regression on console read

In reply to Sasha Levin

From: Jouni Malinen <[email protected]>

[ Upstream commit bebe4681d0e7e1be2608282dc86645728bc7f623 ]

The conversion of UML to use epoll based IRQ controller claimed that clone_one_chan() can safely call um_free_irq() while starting to ignore the delay_free_irq parameter that explicitly noted that the IRQ cannot be freed because this is being called from chan_interrupt(). This resulted in free_irq() getting called in interrupt context ("Trying to free IRQ 6 from IRQ context!").

Fix this by restoring previously used delay_free_irq processing.

Fixes ff6a17989c08 ("Epoll based IRQ controller")
---
 arch/um/drivers/chan_kern.c | 52 +++++++++++++++++++++++++++++++------
 arch/um/kernel/irq.c        |  4 +++
 2 files changed, 48 insertions(+), 8 deletions(-)
arch/um/drivers/chan_kern.c +44 -8
index 05588f9466c7..13ba195f9c9c 100644
@@ -171,19 +171,55 @@ int enable_chan(struct line *line)
 	return err;
 }
 
+/* Items are added in IRQ context, when free_irq can't be called, and
+ * removed in process context, when it can.
+ * This handles interrupt sources which disappear, and which need to
+ * be permanently disabled.  This is discovered in IRQ context, but
+ * the freeing of the IRQ must be done later.
+ */
+static DEFINE_SPINLOCK(irqs_to_free_lock);
+static LIST_HEAD(irqs_to_free);
+
+void free_irqs(void)
+{
+	struct chan *chan;
+	LIST_HEAD(list);
+	struct list_head *ele;
+	unsigned long flags;
+
+	spin_lock_irqsave(&irqs_to_free_lock, flags);
+	list_splice_init(&irqs_to_free, &list);
+	spin_unlock_irqrestore(&irqs_to_free_lock, flags);
+
+	list_for_each(ele, &list) {
+		chan = list_entry(ele, struct chan, free_list);
+
+		if (chan->input && chan->enabled)
+			um_free_irq(chan->line->driver->read_irq, chan);
+		if (chan->output && chan->enabled)
+			um_free_irq(chan->line->driver->write_irq, chan);
+		chan->enabled = 0;
+	}
+}
+
 static void close_one_chan(struct chan *chan, int delay_free_irq)
 {
+	unsigned long flags;
+
 	if (!chan->opened)
 		return;
 
-    /* we can safely call free now - it will be marked
-     *  as free and freed once the IRQ stopped processing
-     */
-	if (chan->input && chan->enabled)
-		um_free_irq(chan->line->driver->read_irq, chan);
-	if (chan->output && chan->enabled)
-		um_free_irq(chan->line->driver->write_irq, chan);
-	chan->enabled = 0;
+	if (delay_free_irq) {
+		spin_lock_irqsave(&irqs_to_free_lock, flags);
+		list_add(&chan->free_list, &irqs_to_free);
+		spin_unlock_irqrestore(&irqs_to_free_lock, flags);
+	} else {
+		if (chan->input && chan->enabled)
+			um_free_irq(chan->line->driver->read_irq, chan);
+		if (chan->output && chan->enabled)
+			um_free_irq(chan->line->driver->write_irq, chan);
+		chan->enabled = 0;
+	}
 	if (chan->ops->close != NULL)
 		(*chan->ops->close)(chan->fd, chan->data);
 
arch/um/kernel/irq.c +4 -0
index 6b7f3827d6e4..2753718d31b9 100644
@@ -21,6 +21,8 @@
 #include <irq_user.h>
 
 
+extern void free_irqs(void);
+
 /* When epoll triggers we do not know why it did so
  * we can also have different IRQs for read and write.
  * This is why we keep a small irq_fd array for each fd -
@@ -100,6 +102,8 @@ void sigio_handler(int sig, struct siginfo *unused_si, struct uml_pt_regs *regs)
 			}
 		}
 	}
+
+	free_irqs();
 }
 
 static int assign_epoll_events_to_irq(struct irq_entry *irq_entry)
Signature · 7 lines
2.20.1


_______________________________________________
linux-um mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-um
author maintainer
#183 lore raw

PATCH 447/671 PM: ACPI/PCI: Resume all devices during hibernation

In reply to Sasha Levin

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit 501debd4aa5edc755037c39ea5a8fba23b41e580 ]

Both the PCI bus type and the ACPI PM domain avoid resuming runtime-suspended devices with DPM_FLAG_SMART_SUSPEND set during hibernation (before creating the snapshot image of system memory), but that turns out to be a mistake. It leads to functional issues and adds complexity that's hard to justify.

For this reason, resume all runtime-suspended PCI devices and all devices in the ACPI PM domains before creating a snapshot image of system memory during hibernation.

Fixes 05087360fd7a (ACPI / PM: Take SMART_SUSPEND driver flag into account)
Fixes c4b65157aeef (PCI / PM: Take SMART_SUSPEND driver flag into account)
---
 drivers/acpi/device_pm.c | 13 +++++++------
 drivers/pci/pci-driver.c | 16 ++++++++--------
 2 files changed, 15 insertions(+), 14 deletions(-)
drivers/acpi/device_pm.c +7 -6
index e0927c5fd282..11b7a1632e5a 100644
@@ -1116,13 +1116,14 @@ EXPORT_SYMBOL_GPL(acpi_subsys_resume_early);
 int acpi_subsys_freeze(struct device *dev)
 {
 	/*
-	 * This used to be done in acpi_subsys_prepare() for all devices and
-	 * some drivers may depend on it, so do it here.  Ideally, however,
-	 * runtime-suspended devices should not be touched during freeze/thaw
-	 * transitions.
+	 * Resume all runtime-suspended devices before creating a snapshot
+	 * image of system memory, because the restore kernel generally cannot
+	 * be expected to always handle them consistently and they need to be
+	 * put into the runtime-active metastate during system resume anyway,
+	 * so it is better to ensure that the state saved in the image will be
+	 * always consistent with that.
 	 */
-	if (!dev_pm_test_driver_flags(dev, DPM_FLAG_SMART_SUSPEND))
-		pm_runtime_resume(dev);
+	pm_runtime_resume(dev);
 
 	return pm_generic_freeze(dev);
 }
drivers/pci/pci-driver.c +8 -8
index e69af9b8361d..5def4b74d54a 100644
@@ -996,15 +996,15 @@ static int pci_pm_freeze(struct device *dev)
 	}
 
 	/*
-	 * This used to be done in pci_pm_prepare() for all devices and some
-	 * drivers may depend on it, so do it here.  Ideally, runtime-suspended
-	 * devices should not be touched during freeze/thaw transitions,
-	 * however.
+	 * Resume all runtime-suspended devices before creating a snapshot
+	 * image of system memory, because the restore kernel generally cannot
+	 * be expected to always handle them consistently and they need to be
+	 * put into the runtime-active metastate during system resume anyway,
+	 * so it is better to ensure that the state saved in the image will be
+	 * always consistent with that.
 	 */
-	if (!dev_pm_smart_suspend_and_suspended(dev)) {
-		pm_runtime_resume(dev);
-		pci_dev->state_saved = false;
-	}
+	pm_runtime_resume(dev);
+	pci_dev->state_saved = false;
 
 	if (pm->freeze) {
 		int error;
Signature · 1 line
2.20.1
author maintainer
#184 lore raw

PATCH 448/671 ACPI: PM: Simplify and fix PM domain hibernation callbacks

In reply to Sasha Levin

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit 3cd7957e85e67120bb9f6bfb75d81dcc19af282b ]

First, after a previous change causing all runtime-suspended devices in the ACPI PM domain (and ACPI LPSS devices) to be resumed before creating a snapshot image of memory during hibernation, it is not necessary to worry about the case in which them might be left in runtime-suspend any more, so get rid of the code related to that from ACPI PM domain and ACPI LPSS hibernation callbacks.

Second, it is not correct to use pm_generic_resume_early() and acpi_subsys_resume_noirq() in hibernation "restore" callbacks (which currently happens in the ACPI PM domain and ACPI LPSS), so introduce proper _restore_late and _restore_noirq callbacks for the ACPI PM domain and ACPI LPSS.

Fixes 05087360fd7a (ACPI / PM: Take SMART_SUSPEND driver flag into account)
---
 drivers/acpi/acpi_lpss.c | 61 ++++++++++++++++++++++++++++++++++------
 drivers/acpi/device_pm.c | 61 ++++++----------------------------------
 include/linux/acpi.h     | 10 -------
 3 files changed, 61 insertions(+), 71 deletions(-)
drivers/acpi/acpi_lpss.c +52 -9
index 30ccd94f87d2..11c460ab9de9 100644
@@ -1086,16 +1086,62 @@ static int acpi_lpss_resume_noirq(struct device *dev)
 	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
 	int ret;
 
-	ret = acpi_subsys_resume_noirq(dev);
+	/* Follow acpi_subsys_resume_noirq(). */
+	if (dev_pm_may_skip_resume(dev))
+		return 0;
+
+	if (dev_pm_smart_suspend_and_suspended(dev))
+		pm_runtime_set_active(dev);
+
+	ret = pm_generic_resume_noirq(dev);
 	if (ret)
 		return ret;
 
-	if (!dev_pm_may_skip_resume(dev) && pdata->dev_desc->resume_from_noirq)
-		ret = acpi_lpss_do_resume_early(dev);
+	if (!pdata->dev_desc->resume_from_noirq)
+		return 0;
 
-	return ret;
+	/*
+	 * The driver's ->resume_early callback will be invoked by
+	 * acpi_lpss_do_resume_early(), with the assumption that the driver
+	 * really wanted to run that code in ->resume_noirq, but it could not
+	 * run before acpi_dev_resume() and the driver expected the latter to be
+	 * called in the "early" phase.
+	 */
+	return acpi_lpss_do_resume_early(dev);
+}
+
+static int acpi_lpss_do_restore_early(struct device *dev)
+{
+	int ret = acpi_lpss_resume(dev);
+
+	return ret ? ret : pm_generic_restore_early(dev);
+}
+
+static int acpi_lpss_restore_early(struct device *dev)
+{
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+
+	if (pdata->dev_desc->resume_from_noirq)
+		return 0;
+
+	return acpi_lpss_do_restore_early(dev);
 }
 
+static int acpi_lpss_restore_noirq(struct device *dev)
+{
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+	int ret;
+
+	ret = pm_generic_restore_noirq(dev);
+	if (ret)
+		return ret;
+
+	if (!pdata->dev_desc->resume_from_noirq)
+		return 0;
+
+	/* This is analogous to what happens in acpi_lpss_resume_noirq(). */
+	return acpi_lpss_do_restore_early(dev);
+}
 #endif /* CONFIG_PM_SLEEP */
 
 static int acpi_lpss_runtime_suspend(struct device *dev)
@@ -1129,14 +1175,11 @@ static struct dev_pm_domain acpi_lpss_pm_domain = {
 		.resume_noirq = acpi_lpss_resume_noirq,
 		.resume_early = acpi_lpss_resume_early,
 		.freeze = acpi_subsys_freeze,
-		.freeze_late = acpi_subsys_freeze_late,
-		.freeze_noirq = acpi_subsys_freeze_noirq,
-		.thaw_noirq = acpi_subsys_thaw_noirq,
 		.poweroff = acpi_subsys_suspend,
 		.poweroff_late = acpi_lpss_suspend_late,
 		.poweroff_noirq = acpi_lpss_suspend_noirq,
-		.restore_noirq = acpi_lpss_resume_noirq,
-		.restore_early = acpi_lpss_resume_early,
+		.restore_noirq = acpi_lpss_restore_noirq,
+		.restore_early = acpi_lpss_restore_early,
 #endif
 		.runtime_suspend = acpi_lpss_runtime_suspend,
 		.runtime_resume = acpi_lpss_runtime_resume,
drivers/acpi/device_pm.c +9 -52

101 lines not on this page · raw message

include/linux/acpi.h +0 -10
index df1252e22dcf..32fabeeda5e3 100644
@@ -917,26 +917,16 @@ int acpi_subsys_prepare(struct device *dev);
 void acpi_subsys_complete(struct device *dev);
 int acpi_subsys_suspend_late(struct device *dev);
 int acpi_subsys_suspend_noirq(struct device *dev);
-int acpi_subsys_resume_noirq(struct device *dev);
-int acpi_subsys_resume_early(struct device *dev);
 int acpi_subsys_suspend(struct device *dev);
 int acpi_subsys_freeze(struct device *dev);
-int acpi_subsys_freeze_late(struct device *dev);
-int acpi_subsys_freeze_noirq(struct device *dev);
-int acpi_subsys_thaw_noirq(struct device *dev);
 #else
 static inline int acpi_dev_resume_early(struct device *dev) { return 0; }
 static inline int acpi_subsys_prepare(struct device *dev) { return 0; }
 static inline void acpi_subsys_complete(struct device *dev) {}
 static inline int acpi_subsys_suspend_late(struct device *dev) { return 0; }
 static inline int acpi_subsys_suspend_noirq(struct device *dev) { return 0; }
-static inline int acpi_subsys_resume_noirq(struct device *dev) { return 0; }
-static inline int acpi_subsys_resume_early(struct device *dev) { return 0; }
 static inline int acpi_subsys_suspend(struct device *dev) { return 0; }
 static inline int acpi_subsys_freeze(struct device *dev) { return 0; }
-static inline int acpi_subsys_freeze_late(struct device *dev) { return 0; }
-static inline int acpi_subsys_freeze_noirq(struct device *dev) { return 0; }
-static inline int acpi_subsys_thaw_noirq(struct device *dev) { return 0; }
 #endif
 
 #ifdef CONFIG_ACPI
Signature · 1 line
2.20.1
author maintainer
#185 lore raw

PATCH 449/671 ACPI: PM: Introduce "poweroff" callbacks for ACPI PM domain and LPSS

In reply to Sasha Levin

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit c95b7595f85c688d5c569ddbbd6ab6a4bdae2f36 ]

In general, it is not correct to call pm_generic_suspend(), pm_generic_suspend_late() and pm_generic_suspend_noirq() during the hibernation's "poweroff" transition, because device drivers may provide special callbacks to be invoked then and the wrappers in question cause system suspend callbacks to be run. Unfortunately, that happens in the ACPI PM domain and ACPI LPSS.

To address this potential issue, introduce "poweroff" callbacks for the ACPI PM and LPSS that will use pm_generic_poweroff(), pm_generic_poweroff_late() and pm_generic_poweroff_noirq() as appropriate.

Fixes 05087360fd7a (ACPI / PM: Take SMART_SUSPEND driver flag into account)
---
 drivers/acpi/acpi_lpss.c | 50 +++++++++++++++++++++++++++++++---
 drivers/acpi/device_pm.c | 58 +++++++++++++++++++++++++++++++++++++---
 include/linux/acpi.h     |  2 ++
 3 files changed, 104 insertions(+), 6 deletions(-)
drivers/acpi/acpi_lpss.c +47 -3
index 11c460ab9de9..ded6c5c17fd7 100644
@@ -1056,6 +1056,13 @@ static int acpi_lpss_suspend_noirq(struct device *dev)
 	int ret;
 
 	if (pdata->dev_desc->resume_from_noirq) {
+		/*
+		 * The driver's ->suspend_late callback will be invoked by
+		 * acpi_lpss_do_suspend_late(), with the assumption that the
+		 * driver really wanted to run that code in ->suspend_noirq, but
+		 * it could not run after acpi_dev_suspend() and the driver
+		 * expected the latter to be called in the "late" phase.
+		 */
 		ret = acpi_lpss_do_suspend_late(dev);
 		if (ret)
 			return ret;
@@ -1142,6 +1149,43 @@ static int acpi_lpss_restore_noirq(struct device *dev)
 	/* This is analogous to what happens in acpi_lpss_resume_noirq(). */
 	return acpi_lpss_do_restore_early(dev);
 }
+
+static int acpi_lpss_do_poweroff_late(struct device *dev)
+{
+	int ret = pm_generic_poweroff_late(dev);
+
+	return ret ? ret : acpi_lpss_suspend(dev, device_may_wakeup(dev));
+}
+
+static int acpi_lpss_poweroff_late(struct device *dev)
+{
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+
+	if (dev_pm_smart_suspend_and_suspended(dev))
+		return 0;
+
+	if (pdata->dev_desc->resume_from_noirq)
+		return 0;
+
+	return acpi_lpss_do_poweroff_late(dev);
+}
+
+static int acpi_lpss_poweroff_noirq(struct device *dev)
+{
+	struct lpss_private_data *pdata = acpi_driver_data(ACPI_COMPANION(dev));
+
+	if (dev_pm_smart_suspend_and_suspended(dev))
+		return 0;
+
+	if (pdata->dev_desc->resume_from_noirq) {
+		/* This is analogous to the acpi_lpss_suspend_noirq() case. */
+		int ret = acpi_lpss_do_poweroff_late(dev);
+		if (ret)
+			return ret;
+	}
+
+	return pm_generic_poweroff_noirq(dev);
+}
 #endif /* CONFIG_PM_SLEEP */
 
 static int acpi_lpss_runtime_suspend(struct device *dev)
@@ -1175,9 +1219,9 @@ static struct dev_pm_domain acpi_lpss_pm_domain = {
 		.resume_noirq = acpi_lpss_resume_noirq,
 		.resume_early = acpi_lpss_resume_early,
 		.freeze = acpi_subsys_freeze,
-		.poweroff = acpi_subsys_suspend,
-		.poweroff_late = acpi_lpss_suspend_late,
-		.poweroff_noirq = acpi_lpss_suspend_noirq,
+		.poweroff = acpi_subsys_poweroff,
+		.poweroff_late = acpi_lpss_poweroff_late,
+		.poweroff_noirq = acpi_lpss_poweroff_noirq,
 		.restore_noirq = acpi_lpss_restore_noirq,
 		.restore_early = acpi_lpss_restore_early,
 #endif
drivers/acpi/device_pm.c +55 -3

70 lines not on this page · raw message

include/linux/acpi.h +2 -0
index 32fabeeda5e3..cd412817654f 100644
@@ -919,6 +919,7 @@ int acpi_subsys_suspend_late(struct device *dev);
 int acpi_subsys_suspend_noirq(struct device *dev);
 int acpi_subsys_suspend(struct device *dev);
 int acpi_subsys_freeze(struct device *dev);
+int acpi_subsys_poweroff(struct device *dev);
 #else
 static inline int acpi_dev_resume_early(struct device *dev) { return 0; }
 static inline int acpi_subsys_prepare(struct device *dev) { return 0; }
@@ -927,6 +928,7 @@ static inline int acpi_subsys_suspend_late(struct device *dev) { return 0; }
 static inline int acpi_subsys_suspend_noirq(struct device *dev) { return 0; }
 static inline int acpi_subsys_suspend(struct device *dev) { return 0; }
 static inline int acpi_subsys_freeze(struct device *dev) { return 0; }
+static inline int acpi_subsys_poweroff(struct device *dev) { return 0; }
 #endif
 
 #ifdef CONFIG_ACPI
Signature · 1 line
2.20.1
author maintainer
#186 lore raw

PATCH 450/671 fsi/core: Fix error paths on CFAM init

In reply to Sasha Levin

From: Jeremy Kerr <[email protected]>

[ Upstream commit 371975b0b07520c85098652d561639837a60a905 ]

Change d1dcd67825 re-worked the struct fsi_slave initialisation in fsi_slave_init, but introduced a few inconsitencies: the slave->dev is now registered through cdev_device_add, but we may kfree() the device out from underneath the cdev registration. We may also leave an IDA allocated.

This change fixes the error paths, so that we kfree() only before the device is registered with the core code. We also move the smode write to before we start creating proper devices, as it's the most likely to fail. We also remove the IDA-allocated minor on error, and properly clean up the of_node.

Fixes d1dcd6782576 ("fsi: Add cfam char devices")
---
 drivers/fsi/fsi-core.c | 32 ++++++++++++++++++++------------
 1 file changed, 20 insertions(+), 12 deletions(-)
drivers/fsi/fsi-core.c +20 -12
index c6fa9b393e84..bd62236d3f97 100644
@@ -1060,6 +1060,14 @@ static int fsi_slave_init(struct fsi_master *master, int link, uint8_t id)
 
 	}
 
+	rc = fsi_slave_set_smode(slave);
+	if (rc) {
+		dev_warn(&master->dev,
+				"can't set smode on slave:%02x:%02x %d\n",
+				link, id, rc);
+		goto err_free;
+	}
+
 	/* Allocate a minor in the FSI space */
 	rc = __fsi_get_new_minor(slave, fsi_dev_cfam, &slave->dev.devt,
 				 &slave->cdev_idx);
@@ -1071,17 +1079,14 @@ static int fsi_slave_init(struct fsi_master *master, int link, uint8_t id)
 	rc = cdev_device_add(&slave->cdev, &slave->dev);
 	if (rc) {
 		dev_err(&slave->dev, "Error %d creating slave device\n", rc);
-		goto err_free;
+		goto err_free_ida;
 	}
 
-	rc = fsi_slave_set_smode(slave);
-	if (rc) {
-		dev_warn(&master->dev,
-				"can't set smode on slave:%02x:%02x %d\n",
-				link, id, rc);
-		kfree(slave);
-		return -ENODEV;
-	}
+	/* Now that we have the cdev registered with the core, any fatal
+	 * failures beyond this point will need to clean up through
+	 * cdev_device_del(). Fortunately though, nothing past here is fatal.
+	 */
+
 	if (master->link_config)
 		master->link_config(master, link,
 				    slave->t_send_delay,
@@ -1098,10 +1103,13 @@ static int fsi_slave_init(struct fsi_master *master, int link, uint8_t id)
 		dev_dbg(&master->dev, "failed during slave scan with: %d\n",
 				rc);
 
-	return rc;
+	return 0;
 
- err_free:
-	put_device(&slave->dev);
+err_free_ida:
+	fsi_free_minor(slave->dev.devt);
+err_free:
+	of_node_put(slave->dev.of_node);
+	kfree(slave);
 	return rc;
 }
 
Signature · 1 line
2.20.1
author maintainer
#187 lore raw

PATCH 451/671 devres: allow const resource arguments

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 9dea44c91469512d346e638694c22c30a5273992 ]

devm_ioremap_resource() does not currently take 'const' arguments, which results in a warning from the first driver trying to do it anyway:

drivers/gpio/gpio-amd-fch.c: In function 'amd_fch_gpio_probe':
drivers/gpio/gpio-amd-fch.c:171:49: error: passing argument 2 of 'devm_ioremap_resource' discards 'const' qualifier from pointer target type [-Werror=discarded-qualifiers]

  priv->base = devm_ioremap_resource(&pdev->dev, &amd_fch_gpio_iores);
                                                 ^~~~~~~~~~~~~~~~~~~

Change the prototype to allow it, as there is no real reason not to.

---
 include/linux/device.h | 3 ++-
 lib/devres.c           | 3 ++-
 2 files changed, 4 insertions(+), 2 deletions(-)
include/linux/device.h +2 -1
index e9d1c768f972..c74ce473589a 100644
@@ -701,7 +701,8 @@ extern unsigned long devm_get_free_pages(struct device *dev,
 					 gfp_t gfp_mask, unsigned int order);
 extern void devm_free_pages(struct device *dev, unsigned long addr);
 
-void __iomem *devm_ioremap_resource(struct device *dev, struct resource *res);
+void __iomem *devm_ioremap_resource(struct device *dev,
+				    const struct resource *res);
 
 void __iomem *devm_of_iomap(struct device *dev,
 			    struct device_node *node, int index,
lib/devres.c +2 -1
index faccf1a037d0..aa0f5308ac6b 100644
@@ -131,7 +131,8 @@ EXPORT_SYMBOL(devm_iounmap);
  *	if (IS_ERR(base))
  *		return PTR_ERR(base);
  */
-void __iomem *devm_ioremap_resource(struct device *dev, struct resource *res)
+void __iomem *devm_ioremap_resource(struct device *dev,
+				    const struct resource *res)
 {
 	resource_size_t size;
 	const char *name;
Signature · 1 line
2.20.1
author maintainer
#188 lore raw

PATCH 452/671 fsi: sbefifo: Don't fail operations when in SBE IPL state

In reply to Sasha Levin

From: Eddie James <[email protected]>

[ Upstream commit 7ce98fb6c53d2311b3e9faae90b1a9c1a96534db ]

SBE fifo operations should be allowed while the SBE is in any of the "IPL" states. Operations should succeed in this state.

---
 drivers/fsi/fsi-sbefifo.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/fsi/fsi-sbefifo.c +2 -2
index ae861342626e..9fa3959e0855 100644
@@ -289,11 +289,11 @@ static int sbefifo_check_sbe_state(struct sbefifo *sbefifo)
 	switch ((sbm & CFAM_SBM_SBE_STATE_MASK) >> CFAM_SBM_SBE_STATE_SHIFT) {
 	case SBE_STATE_UNKNOWN:
 		return -ESHUTDOWN;
+	case SBE_STATE_DMT:
+		return -EBUSY;
 	case SBE_STATE_IPLING:
 	case SBE_STATE_ISTEP:
 	case SBE_STATE_MPIPL:
-	case SBE_STATE_DMT:
-		return -EBUSY;
 	case SBE_STATE_RUNTIME:
 	case SBE_STATE_DUMP: /* Not sure about that one */
 		break;
Signature · 1 line
2.20.1
author maintainer
#189 lore raw

PATCH 453/671 RDMA/hns: Fixs hw access invalid dma memory error

In reply to Sasha Levin

From: Xi Wang <[email protected]>

[ Upstream commit ec5bc2cc69b4fc494e04d10fc5226f6f9cf67c56 ]

When smmu is enable, if execute the perftest command and then use 'kill -9' to exit, follow this operation repeatedly, the kernel will have a high probability to print the following smmu event:

arm-smmu-v3 arm-smmu-v3.1.auto: event 0x10 received:

  arm-smmu-v3 arm-smmu-v3.1.auto:  0x00007d0000000010
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x0000020900000080
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x00000000f47cf000
  arm-smmu-v3 arm-smmu-v3.1.auto:  0x00000000f47cf000

This is because the hw will periodically refresh the qpc cache until the next reset.

This patch fixed it by removing the action that release qpc memory in the 'hns_roce_qp_free' function.

Fixes 9a4435375cd1 ("IB/hns: Add driver files for hns RoCE driver")
---
 drivers/infiniband/hw/hns/hns_roce_qp.c | 1 -
 1 file changed, 1 deletion(-)
drivers/infiniband/hw/hns/hns_roce_qp.c +0 -1
index af24698ff226..3012d7eb4ccb 100644
@@ -262,7 +262,6 @@ void hns_roce_qp_free(struct hns_roce_dev *hr_dev, struct hns_roce_qp *hr_qp)
 			hns_roce_table_put(hr_dev, &qp_table->trrl_table,
 					   hr_qp->qpn);
 		hns_roce_table_put(hr_dev, &qp_table->irrl_table, hr_qp->qpn);
-		hns_roce_table_put(hr_dev, &qp_table->qp_table, hr_qp->qpn);
 	}
 }
 EXPORT_SYMBOL_GPL(hns_roce_qp_free);
Signature · 1 line
2.20.1
author maintainer
#190 lore raw

PATCH 454/671 PCI: mobiveil: Remove the flag MSI_FLAG_MULTI_PCI_MSI

In reply to Sasha Levin

From: Hou Zhiqiang <[email protected]>

[ Upstream commit a131fb6364c1be0924dcb969ecf6b988c556a5d5 ]

The Mobiveil internal MSI controller requires separate target addresses, one per MSI vector; this is clearly incompatible with the Multiple MSI feature, which requires the same target address for all vectors requested by an endpoint (ie the Message Address field in the MSI Capability structure), so the multi MSI feature is clearly not supported by the host controller driver.

Remove the flag MSI_FLAG_MULTI_PCI_MSI and with it multi MSI support, fixing the misconfiguration.

Fixes 1e913e58335f ("PCI: mobiveil: Add MSI support")
[[email protected]: commit log]
---
 drivers/pci/controller/pcie-mobiveil.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/controller/pcie-mobiveil.c +1 -1
index a2d1e89d4867..dc228eb500ed 100644
@@ -643,7 +643,7 @@ static struct irq_chip mobiveil_msi_irq_chip = {
 
 static struct msi_domain_info mobiveil_msi_domain_info = {
 	.flags	= (MSI_FLAG_USE_DEF_DOM_OPS | MSI_FLAG_USE_DEF_CHIP_OPS |
-		MSI_FLAG_MULTI_PCI_MSI | MSI_FLAG_PCI_MSIX),
+		   MSI_FLAG_PCI_MSIX),
 	.chip	= &mobiveil_msi_irq_chip,
 };
 
Signature · 1 line
2.20.1
author maintainer
#191 lore raw

PATCH 455/671 PCI: mobiveil: Fix devfn check in mobiveil_pcie_valid_device()

In reply to Sasha Levin

From: Hou Zhiqiang <[email protected]>

[ Upstream commit cbd50b3ca3964c79dac65fda277637577e029e8c ]

Current check for devfn number in mobiveil_pci_valid_device() is wrong in that it flags as invalid functions present in PCI device 0 in the root bus while it is perfectly valid to access all functions in PCI device 0 in the root bus.

Update the check in mobiveil_pci_valid_device() to fix the issue.

Fixes 9af6bcb11e12 ("PCI: mobiveil: Add Mobiveil PCIe Host Bridge IP driver")
---
 drivers/pci/controller/pcie-mobiveil.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/pci/controller/pcie-mobiveil.c +1 -1
index dc228eb500ed..476be4f3c7f6 100644
@@ -174,7 +174,7 @@ static bool mobiveil_pcie_valid_device(struct pci_bus *bus, unsigned int devfn)
 	 * Do not read more than one device on the bus directly
 	 * attached to RC
 	 */
-	if ((bus->primary == pcie->root_bus_nr) && (devfn > 0))
+	if ((bus->primary == pcie->root_bus_nr) && (PCI_SLOT(devfn) > 0))
 		return false;
 
 	return true;
Signature · 1 line
2.20.1
author maintainer
#192 lore raw

PATCH 456/671 PCI: mobiveil: Fix the valid check for inbound and outbound windows

In reply to Sasha Levin

From: Hou Zhiqiang <[email protected]>

[ Upstream commit ccd34dac2ed596b1f26079912bdf638e002a3979 ]

In program_ib/ob_windows() check the window index from the function parameter instead of the total number of initialized windows to determine if the specified window is valid.

Fixes 9af6bcb11e12 ("PCI: mobiveil: Add Mobiveil PCIe Host Bridge IP driver")
---
 drivers/pci/controller/pcie-mobiveil.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/pci/controller/pcie-mobiveil.c +2 -2
index 476be4f3c7f6..14f816591e84 100644
@@ -395,7 +395,7 @@ static void program_ib_windows(struct mobiveil_pcie *pcie, int win_num,
 	int amap_ctrl_dw;
 	u64 size64 = ~(size - 1);
 
-	if ((pcie->ib_wins_configured + 1) > pcie->ppio_wins) {
+	if (win_num >= pcie->ppio_wins) {
 		dev_err(&pcie->pdev->dev,
 			"ERROR: max inbound windows reached !\n");
 		return;
@@ -429,7 +429,7 @@ static void program_ob_windows(struct mobiveil_pcie *pcie, int win_num,
 	u32 value, type;
 	u64 size64 = ~(size - 1);
 
-	if ((pcie->ob_wins_configured + 1) > pcie->apio_wins) {
+	if (win_num >= pcie->apio_wins) {
 		dev_err(&pcie->pdev->dev,
 			"ERROR: max outbound windows reached !\n");
 		return;
Signature · 1 line
2.20.1
author maintainer
#193 lore raw

PATCH 457/671 ceph: fix "ceph.dir.rctime" vxattr value

In reply to Sasha Levin

From: David Disseldorp <[email protected]>

[ Upstream commit 718807289d4130be1fe13f24f018733116958070 ]

The vxattr value incorrectly places a "09" prefix to the nanoseconds field, instead of providing it as a zero-pad width specifier after '%'.

Fixes 3489b42a72a4 ("ceph: fix three bugs, two in ceph_vxattrcb_file_layout()")
---
 fs/ceph/xattr.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/ceph/xattr.c +1 -1
index 5e4f3f833e85..a09ce27ab220 100644
@@ -221,7 +221,7 @@ static size_t ceph_vxattrcb_dir_rbytes(struct ceph_inode_info *ci, char *val,
 static size_t ceph_vxattrcb_dir_rctime(struct ceph_inode_info *ci, char *val,
 				       size_t size)
 {
-	return snprintf(val, size, "%lld.09%ld", ci->i_rctime.tv_sec,
+	return snprintf(val, size, "%lld.%09ld", ci->i_rctime.tv_sec,
 			ci->i_rctime.tv_nsec);
 }
 
Signature · 1 line
2.20.1
author maintainer
#194 lore raw

PATCH 458/671 net: pasemi: fix an use-after-free in pasemi_mac_phy_init()

In reply to Sasha Levin

From: Wen Yang <[email protected]>

[ Upstream commit faf5577f2498cea23011b5c785ef853ded22700b ]

The phy_dn variable is still being used in of_phy_connect() after the of_node_put() call, which may result in use-after-free.

Fixes 1dd2d06c0459 ("net: Rework pasemi_mac driver to use of_mdio infrastructure")
Cc "David S. Miller" <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Luis Chamberlain <[email protected]>
Cc Michael Ellerman <[email protected]>
---
 drivers/net/ethernet/pasemi/pasemi_mac.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/pasemi/pasemi_mac.c +1 -1
index 8a31a02c9f47..65f69e562618 100644
@@ -1053,7 +1053,6 @@ static int pasemi_mac_phy_init(struct net_device *dev)
 
 	dn = pci_device_to_OF_node(mac->pdev);
 	phy_dn = of_parse_phandle(dn, "phy-handle", 0);
-	of_node_put(phy_dn);
 
 	mac->link = 0;
 	mac->speed = 0;
@@ -1062,6 +1061,7 @@ static int pasemi_mac_phy_init(struct net_device *dev)
 	phydev = of_phy_connect(dev, phy_dn, &pasemi_adjust_link, 0,
 				PHY_INTERFACE_MODE_SGMII);
 
+	of_node_put(phy_dn);
 	if (!phydev) {
 		printk(KERN_ERR "%s: Could not attach to phy\n", dev->name);
 		return -ENODEV;
Signature · 1 line
2.20.1
author maintainer
#195 lore raw

PATCH 459/671 net/tls: fix socket wmem accounting on fallback with netem

In reply to Sasha Levin

From: Jakub Kicinski <[email protected]>

[ Upstream commit 5c4b4608fe100838c62591877101128467e56c00 ]

netem runs skb_orphan_partial() which "disconnects" the skb from normal TCP write memory accounting. We should not adjust sk->sk_wmem_alloc on the fallback path for such skbs.

Fixes e8f69799810c ("net/tls: Add generic NIC offload infrastructure")
---
 net/tls/tls_device_fallback.c | 4 ++++
 1 file changed, 4 insertions(+)
net/tls/tls_device_fallback.c +4 -0
index 426dd97725e4..6cf832891b53 100644
@@ -208,6 +208,10 @@ static void complete_skb(struct sk_buff *nskb, struct sk_buff *skb, int headln)
 
 	update_chksum(nskb, headln);
 
+	/* sock_efree means skb must gone through skb_orphan_partial() */
+	if (nskb->destructor == sock_efree)
+		return;
+
 	delta = nskb->truesize - skb->truesize;
 	if (likely(delta < 0))
 		WARN_ON_ONCE(refcount_sub_and_test(-delta, &sk->sk_wmem_alloc));
Signature · 1 line
2.20.1
author maintainer
#196 lore raw

PATCH 460/671 x86/pgtable/32: Fix LOWMEM_PAGES constant

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 26515699863d68058e290e18e83f444925920be5 ]

clang points out that the computation of LOWMEM_PAGES causes a signed integer overflow on 32-bit x86:

arch/x86/kernel/head32.c:83:20: error: signed shift result (0x100000000) requires 34 bits to represent, but 'int' only has 32 bits [-Werror,-Wshift-overflow]

                (PAGE_TABLE_SIZE(LOWMEM_PAGES) << PAGE_SHIFT);
                                 ^~~~~~~~~~~~
arch/x86/include/asm/pgtable_32.h:109:27: note: expanded from macro 'LOWMEM_PAGES'
 #define LOWMEM_PAGES ((((2<<31) - __PAGE_OFFSET) >> PAGE_SHIFT))
                         ~^ ~~

arch/x86/include/asm/pgtable_32.h:98:34: note: expanded from macro 'PAGE_TABLE_SIZE'
#define PAGE_TABLE_SIZE(pages) ((pages) / PTRS_PER_PGD)

Use the _ULL() macro to make it a 64-bit constant.

Fixes 1e620f9b23e5 ("x86/boot/32: Convert the 32-bit pgtable setup code from assembly to C")
---
 arch/x86/include/asm/pgtable_32.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/include/asm/pgtable_32.h +1 -1
index b3ec519e3982..71e1df860176 100644
@@ -106,6 +106,6 @@ do {						\
  * with only a host target support using a 32-bit type for internal
  * representation.
  */
-#define LOWMEM_PAGES ((((2<<31) - __PAGE_OFFSET) >> PAGE_SHIFT))
+#define LOWMEM_PAGES ((((_ULL(2)<<31) - __PAGE_OFFSET) >> PAGE_SHIFT))
 
 #endif /* _ASM_X86_PGTABLE_32_H */
Signature · 1 line
2.20.1
author maintainer
#197 lore raw

PATCH 461/671 xdp: fix possible cq entry leak

In reply to Sasha Levin

From: Ilya Maximets <[email protected]>

[ Upstream commit 675716400da6f15b9d3db04ef74ee74ca9a00af3 ]

Completion queue address reservation could not be undone. In case of bad 'queue_id' or skb allocation failure, reserved entry will be leaked reducing the total capacity of completion queue.

Fix that by moving reservation to the point where failure is not possible. Additionally, 'queue_id' checking moved out from the loop since there is no point to check it there.

---
 net/xdp/xsk.c | 11 ++++-------
 1 file changed, 4 insertions(+), 7 deletions(-)
net/xdp/xsk.c +4 -7
index 547fc4554b22..c90854bc3048 100644
@@ -218,6 +218,9 @@ static int xsk_generic_xmit(struct sock *sk, struct msghdr *m,
 
 	mutex_lock(&xs->mutex);
 
+	if (xs->queue_id >= xs->dev->real_num_tx_queues)
+		goto out;
+
 	while (xskq_peek_desc(xs->tx, &desc)) {
 		char *buffer;
 		u64 addr;
@@ -228,12 +231,6 @@ static int xsk_generic_xmit(struct sock *sk, struct msghdr *m,
 			goto out;
 		}
 
-		if (xskq_reserve_addr(xs->umem->cq))
-			goto out;
-
-		if (xs->queue_id >= xs->dev->real_num_tx_queues)
-			goto out;
-
 		len = desc.len;
 		skb = sock_alloc_send_skb(sk, len, 1, &err);
 		if (unlikely(!skb)) {
@@ -245,7 +242,7 @@ static int xsk_generic_xmit(struct sock *sk, struct msghdr *m,
 		addr = desc.addr;
 		buffer = xdp_umem_get_data(xs->umem, addr);
 		err = skb_store_bits(skb, 0, buffer, len);
-		if (unlikely(err)) {
+		if (unlikely(err) || xskq_reserve_addr(xs->umem->cq)) {
 			kfree_skb(skb);
 			goto out;
 		}
Signature · 1 line
2.20.1
author maintainer
#198 lore raw

PATCH 462/671 ARM: stm32: use "depends on" instead of "if" after prompt

In reply to Sasha Levin

From: Masahiro Yamada <[email protected]>

[ Upstream commit 7e8a0f10899075ac2665c78c4e49dbaf32bf3346 ]

This appeared after the global fixups by commit e32465429490 ("ARM: use "depends on" for SoC configs instead of "if" after prompt"). Fix it now.

Fixes e32465429490 ("ARM: use "depends on" for SoC configs instead of "if" after prompt")
---
 arch/arm/mach-stm32/Kconfig | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
arch/arm/mach-stm32/Kconfig +2 -1
index 713c068b953f..adca4368d67c 100644
@@ -1,5 +1,6 @@
 menuconfig ARCH_STM32
-	bool "STMicroelectronics STM32 family" if ARM_SINGLE_ARMV7M || ARCH_MULTI_V7
+	bool "STMicroelectronics STM32 family"
+	depends on ARM_SINGLE_ARMV7M || ARCH_MULTI_V7
 	select ARMV7M_SYSTICK if ARM_SINGLE_ARMV7M
 	select HAVE_ARM_ARCH_TIMER if ARCH_MULTI_V7
 	select ARM_GIC if ARCH_MULTI_V7
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#199 lore raw

PATCH 463/671 scsi: libfc: fix null pointer dereference on a null lport

In reply to Sasha Levin

From: Colin Ian King <[email protected]>

[ Upstream commit 41a6bf6529edd10a6def42e3b2c34a7474bcc2f5 ]

Currently if lport is null then the null lport pointer is dereference when printing out debug via the FC_LPORT_DB macro. Fix this by using the more generic FC_LIBFC_DBG debug macro instead that does not use lport.

Addresses-Coverity: ("Dereference after null check")

Fixes 7414705ea4ae ("libfc: Add runtime debugging with debug_logging module parameter")
---
 drivers/scsi/libfc/fc_exch.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/scsi/libfc/fc_exch.c +1 -1
index 42bcf7f3a0f9..6ba257cbc6d9 100644
@@ -2603,7 +2603,7 @@ void fc_exch_recv(struct fc_lport *lport, struct fc_frame *fp)
 
 	/* lport lock ? */
 	if (!lport || lport->state == LPORT_ST_DISABLED) {
-		FC_LPORT_DBG(lport, "Receiving frames for an lport that "
+		FC_LIBFC_DBG("Receiving frames for an lport that "
 			     "has not been initialized correctly\n");
 		fc_frame_free(fp);
 		return;
Signature · 1 line
2.20.1
author maintainer
#200 lore raw

PATCH 464/671 xfrm interface: ifname may be wrong in logs

In reply to Sasha Levin

From: Nicolas Dichtel <[email protected]>

[ Upstream commit e0aaa332e6a97dae57ad59cdb19e21f83c3d081c ]

The ifname is copied when the interface is created, but is never updated later. In fact, this property is used only in one error message, where the netdevice pointer is available, thus let's use it.

Fixes f203b76d7809 ("xfrm: Add virtual xfrm interfaces")
---
 include/net/xfrm.h        |  1 -
 net/xfrm/xfrm_interface.c | 10 +---------
 2 files changed, 1 insertion(+), 10 deletions(-)
include/net/xfrm.h +0 -1
index fb9b19a3b749..48dc1ce2170d 100644
@@ -1054,7 +1054,6 @@ static inline void xfrm_dst_destroy(struct xfrm_dst *xdst)
 void xfrm_dst_ifdown(struct dst_entry *dst, struct net_device *dev);
 
 struct xfrm_if_parms {
-	char name[IFNAMSIZ];	/* name of XFRM device */
 	int link;		/* ifindex of underlying L2 interface */
 	u32 if_id;		/* interface identifyer */
 };
net/xfrm/xfrm_interface.c +1 -9
index d6a3cdf7885c..4ee512622e93 100644
@@ -145,8 +145,6 @@ static int xfrmi_create(struct net_device *dev)
 	if (err < 0)
 		goto out;
 
-	strcpy(xi->p.name, dev->name);
-
 	dev_hold(dev);
 	xfrmi_link(xfrmn, xi);
 
@@ -293,7 +291,7 @@ xfrmi_xmit2(struct sk_buff *skb, struct net_device *dev, struct flowi *fl)
 	if (tdev == dev) {
 		stats->collisions++;
 		net_warn_ratelimited("%s: Local routing loop detected!\n",
-				     xi->p.name);
+				     dev->name);
 		goto tx_err_dst_release;
 	}
 
@@ -648,12 +646,6 @@ static int xfrmi_newlink(struct net *src_net, struct net_device *dev,
 	int err;
 
 	xfrmi_netlink_parms(data, &p);
-
-	if (!tb[IFLA_IFNAME])
-		return -EINVAL;
-
-	nla_strlcpy(p.name, tb[IFLA_IFNAME], IFNAMSIZ);
-
 	xi = xfrmi_locate(net, &p);
 	if (xi)
 		return -EEXIST;
Signature · 1 line
2.20.1
author maintainer
#201 lore raw

PATCH 465/671 drm/panel: make drm_panel.h self-contained

In reply to Sasha Levin

From: Jani Nikula <[email protected]>

[ Upstream commit bf3f5e98559360661a3d2af340d46522512c0b00 ]

Fix build warning if drm_panel.h is built with CONFIG_OF=n or CONFIG_DRM_PANEL=n and included without the prerequisite err.h:

./include/drm/drm_panel.h: In function ‘of_drm_find_panel’:
./include/drm/drm_panel.h:203:9: error: implicit declaration of function ‘ERR_PTR’ [-Werror=implicit-function-declaration]

  return ERR_PTR(-ENODEV);
         ^~~~~~~
./include/drm/drm_panel.h:203:9: error: returning ‘int’ from a function with return type ‘struct drm_panel *’ makes pointer from integer without a cast [-Werror=int-conversion]
  return ERR_PTR(-ENODEV);
         ^~~~~~~~~~~~~~~~
Fixes 5fa8e4a22182 ("drm/panel: Make of_drm_find_panel() return an ERR_PTR() instead of NULL")
Cc Boris Brezillon <[email protected]>
---
 include/drm/drm_panel.h | 1 +
 1 file changed, 1 insertion(+)
include/drm/drm_panel.h +1 -0
index 777814755fa6..675aa1e876ce 100644
@@ -24,6 +24,7 @@
 #ifndef __DRM_PANEL_H__
 #define __DRM_PANEL_H__
 
+#include <linux/err.h>
 #include <linux/errno.h>
 #include <linux/list.h>
 
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author maintainer
#202 lore raw

PATCH 466/671 clk: sunxi-ng: v3s: add the missing PLL_DDR1

In reply to Sasha Levin

From: Icenowy Zheng <[email protected]>

[ Upstream commit c5ed9475c22c89d5409402055142372e35d26a3f ]

The user manual of V3/V3s/S3 declares a PLL_DDR1, however it's forgot when developing the V3s CCU driver.

Add back the missing PLL_DDR1.

Fixes d0f11d14b0bc ("clk: sunxi-ng: add support for V3s CCU")
---
 drivers/clk/sunxi-ng/ccu-sun8i-v3s.c | 19 +++++++++++++++----
 drivers/clk/sunxi-ng/ccu-sun8i-v3s.h |  6 ++++--
 2 files changed, 19 insertions(+), 6 deletions(-)
drivers/clk/sunxi-ng/ccu-sun8i-v3s.c +15 -4
index 9e3f4088724b..c7f9d974b10d 100644
@@ -84,7 +84,7 @@ static SUNXI_CCU_NM_WITH_FRAC_GATE_LOCK(pll_ve_clk, "pll-ve",
 					BIT(28),	/* lock */
 					0);
 
-static SUNXI_CCU_NKM_WITH_GATE_LOCK(pll_ddr_clk, "pll-ddr",
+static SUNXI_CCU_NKM_WITH_GATE_LOCK(pll_ddr0_clk, "pll-ddr0",
 				    "osc24M", 0x020,
 				    8, 5,	/* N */
 				    4, 2,	/* K */
@@ -123,6 +123,14 @@ static SUNXI_CCU_NK_WITH_GATE_LOCK_POSTDIV(pll_periph1_clk, "pll-periph1",
 					   2,		/* post-div */
 					   0);
 
+static SUNXI_CCU_NM_WITH_GATE_LOCK(pll_ddr1_clk, "pll-ddr1",
+				   "osc24M", 0x04c,
+				   8, 7,	/* N */
+				   0, 2,	/* M */
+				   BIT(31),	/* gate */
+				   BIT(28),	/* lock */
+				   0);
+
 static const char * const cpu_parents[] = { "osc32k", "osc24M",
 					     "pll-cpu", "pll-cpu" };
 static SUNXI_CCU_MUX(cpu_clk, "cpu", cpu_parents,
@@ -310,7 +318,8 @@ static SUNXI_CCU_GATE(usb_phy0_clk,	"usb-phy0",	"osc24M",
 static SUNXI_CCU_GATE(usb_ohci0_clk,	"usb-ohci0",	"osc24M",
 		      0x0cc, BIT(16), 0);
 
-static const char * const dram_parents[] = { "pll-ddr", "pll-periph0-2x" };
+static const char * const dram_parents[] = { "pll-ddr0", "pll-ddr1",
+					     "pll-periph0-2x" };
 static SUNXI_CCU_M_WITH_MUX(dram_clk, "dram", dram_parents,
 			    0x0f4, 0, 4, 20, 2, CLK_IS_CRITICAL);
 
@@ -369,10 +378,11 @@ static struct ccu_common *sun8i_v3s_ccu_clks[] = {
 	&pll_audio_base_clk.common,
 	&pll_video_clk.common,
 	&pll_ve_clk.common,
-	&pll_ddr_clk.common,
+	&pll_ddr0_clk.common,
 	&pll_periph0_clk.common,
 	&pll_isp_clk.common,
 	&pll_periph1_clk.common,
+	&pll_ddr1_clk.common,
 	&cpu_clk.common,
 	&axi_clk.common,
 	&ahb1_clk.common,
@@ -457,11 +467,12 @@ static struct clk_hw_onecell_data sun8i_v3s_hw_clks = {
 		[CLK_PLL_AUDIO_8X]	= &pll_audio_8x_clk.hw,
 		[CLK_PLL_VIDEO]		= &pll_video_clk.common.hw,
 		[CLK_PLL_VE]		= &pll_ve_clk.common.hw,
-		[CLK_PLL_DDR]		= &pll_ddr_clk.common.hw,
+		[CLK_PLL_DDR0]		= &pll_ddr0_clk.common.hw,
 		[CLK_PLL_PERIPH0]	= &pll_periph0_clk.common.hw,
 		[CLK_PLL_PERIPH0_2X]	= &pll_periph0_2x_clk.hw,
 		[CLK_PLL_ISP]		= &pll_isp_clk.common.hw,
 		[CLK_PLL_PERIPH1]	= &pll_periph1_clk.common.hw,
+		[CLK_PLL_DDR1]		= &pll_ddr1_clk.common.hw,
 		[CLK_CPU]		= &cpu_clk.common.hw,
 		[CLK_AXI]		= &axi_clk.common.hw,
 		[CLK_AHB1]		= &ahb1_clk.common.hw,
drivers/clk/sunxi-ng/ccu-sun8i-v3s.h +4 -2
index 4a4d36fdad96..a091b7217dfd 100644
@@ -29,7 +29,7 @@
 #define CLK_PLL_AUDIO_8X	5
 #define CLK_PLL_VIDEO		6
 #define CLK_PLL_VE		7
-#define CLK_PLL_DDR		8
+#define CLK_PLL_DDR0		8
 #define CLK_PLL_PERIPH0		9
 #define CLK_PLL_PERIPH0_2X	10
 #define CLK_PLL_ISP		11
@@ -58,6 +58,8 @@
 
 /* And the GPU module clock is exported */
 
-#define CLK_NUMBER		(CLK_MIPI_CSI + 1)
+#define CLK_PLL_DDR1		74
+
+#define CLK_NUMBER		(CLK_PLL_DDR1 + 1)
 
 #endif /* _CCU_SUN8I_H3_H_ */
Signature · 1 line
2.20.1
author maintainer
#203 lore raw

PATCH 467/671 PM: sleep: Fix possible overflow in pm_system_cancel_wakeup()

In reply to Sasha Levin

From: "Rafael J. Wysocki" <[email protected]>

[ Upstream commit 2933954b71f10d392764f95eec0f0aa2d103054b ]

It is not actually guaranteed that pm_abort_suspend will be nonzero when pm_system_cancel_wakeup() is called which may lead to subtle issues, so make it use atomic_dec_if_positive() instead of atomic_dec() for the safety sake.

Fixes 33e4f80ee69b ("ACPI / PM: Ignore spurious SCI wakeups from suspend-to-idle")
---
 drivers/base/power/wakeup.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/base/power/wakeup.c +1 -1
index 7c84f64c74f7..2dfa2e048745 100644
@@ -875,7 +875,7 @@ EXPORT_SYMBOL_GPL(pm_system_wakeup);
 
 void pm_system_cancel_wakeup(void)
 {
-	atomic_dec(&pm_abort_suspend);
+	atomic_dec_if_positive(&pm_abort_suspend);
 }
 
 void pm_wakeup_clear(bool reset)
Signature · 1 line
2.20.1
author maintainer
#204 lore raw

PATCH 468/671 libertas_tf: Use correct channel range in lbtf_geo_init

In reply to Sasha Levin

From: YueHaibing <[email protected]>

[ Upstream commit 2ec4ad49b98e4a14147d04f914717135eca7c8b1 ]

It seems we should use 'range' instead of 'priv->range' in lbtf_geo_init(), because 'range' is the corret one related to current regioncode.

Fixes 691cdb49388b ("libertas_tf: command helper functions for libertas_tf")
---
 drivers/net/wireless/marvell/libertas_tf/cmd.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/marvell/libertas_tf/cmd.c +1 -1
index 909ac3685010..2b193f1257a5 100644
@@ -69,7 +69,7 @@ static void lbtf_geo_init(struct lbtf_private *priv)
 			break;
 		}
 
-	for (ch = priv->range.start; ch < priv->range.end; ch++)
+	for (ch = range->start; ch < range->end; ch++)
 		priv->channels[CHAN_TO_IDX(ch)].flags = 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#205 lore raw

PATCH 469/671 qed: reduce maximum stack frame size

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 7c116e02a4a7575c8c62bfd2106e3e3ec8fb99dc ]

clang warns about an overly large stack frame in one function when it decides to inline all __qed_get_vport_*() functions into __qed_get_vport_stats():

drivers/net/ethernet/qlogic/qed/qed_l2.c:1889:13: error: stack frame size of 1128 bytes in function '_qed_get_vport_stats' [-Werror,-Wframe-larger-than=]

Use a noinline_for_stack annotation to prevent clang from inlining these, which keeps the maximum stack usage at around half of that in the worst case, similar to what we get with gcc.

Fixes 86622ee75312 ("qed: Move statistics to L2 code")
---
 drivers/net/ethernet/qlogic/qed/qed_l2.c | 34 +++++++++++-------------
 1 file changed, 15 insertions(+), 19 deletions(-)
drivers/net/ethernet/qlogic/qed/qed_l2.c +15 -19
index 64ac95ca4df2..d921b991dbdb 100644
@@ -1631,10 +1631,9 @@ static void __qed_get_vport_pstats_addrlen(struct qed_hwfn *p_hwfn,
 	}
 }
 
-static void __qed_get_vport_pstats(struct qed_hwfn *p_hwfn,
-				   struct qed_ptt *p_ptt,
-				   struct qed_eth_stats *p_stats,
-				   u16 statistics_bin)
+static noinline_for_stack void
+__qed_get_vport_pstats(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt,
+		       struct qed_eth_stats *p_stats, u16 statistics_bin)
 {
 	struct eth_pstorm_per_queue_stat pstats;
 	u32 pstats_addr = 0, pstats_len = 0;
@@ -1661,10 +1660,9 @@ static void __qed_get_vport_pstats(struct qed_hwfn *p_hwfn,
 	    HILO_64_REGPAIR(pstats.error_drop_pkts);
 }
 
-static void __qed_get_vport_tstats(struct qed_hwfn *p_hwfn,
-				   struct qed_ptt *p_ptt,
-				   struct qed_eth_stats *p_stats,
-				   u16 statistics_bin)
+static noinline_for_stack void
+__qed_get_vport_tstats(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt,
+		       struct qed_eth_stats *p_stats, u16 statistics_bin)
 {
 	struct tstorm_per_port_stat tstats;
 	u32 tstats_addr, tstats_len;
@@ -1709,10 +1707,9 @@ static void __qed_get_vport_ustats_addrlen(struct qed_hwfn *p_hwfn,
 	}
 }
 
-static void __qed_get_vport_ustats(struct qed_hwfn *p_hwfn,
-				   struct qed_ptt *p_ptt,
-				   struct qed_eth_stats *p_stats,
-				   u16 statistics_bin)
+static noinline_for_stack
+void __qed_get_vport_ustats(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt,
+			    struct qed_eth_stats *p_stats, u16 statistics_bin)
 {
 	struct eth_ustorm_per_queue_stat ustats;
 	u32 ustats_addr = 0, ustats_len = 0;
@@ -1751,10 +1748,9 @@ static void __qed_get_vport_mstats_addrlen(struct qed_hwfn *p_hwfn,
 	}
 }
 
-static void __qed_get_vport_mstats(struct qed_hwfn *p_hwfn,
-				   struct qed_ptt *p_ptt,
-				   struct qed_eth_stats *p_stats,
-				   u16 statistics_bin)
+static noinline_for_stack void
+__qed_get_vport_mstats(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt,
+		       struct qed_eth_stats *p_stats, u16 statistics_bin)
 {
 	struct eth_mstorm_per_queue_stat mstats;
 	u32 mstats_addr = 0, mstats_len = 0;
@@ -1780,9 +1776,9 @@ static void __qed_get_vport_mstats(struct qed_hwfn *p_hwfn,
 	    HILO_64_REGPAIR(mstats.tpa_coalesced_bytes);
 }
 
-static void __qed_get_vport_port_stats(struct qed_hwfn *p_hwfn,
-				       struct qed_ptt *p_ptt,
-				       struct qed_eth_stats *p_stats)
+static noinline_for_stack void
+__qed_get_vport_port_stats(struct qed_hwfn *p_hwfn, struct qed_ptt *p_ptt,
+			   struct qed_eth_stats *p_stats)
 {
 	struct qed_eth_stats_common *p_common = &p_stats->common;
 	struct port_stats port_stats;
Signature · 1 line
2.20.1
author maintainer
#206 lore raw

PATCH 470/671 usb: host: xhci-hub: fix extra endianness conversion

In reply to Sasha Levin

From: Ruslan Bilovol <[email protected]>

[ Upstream commit 6269e4c76eacabaea0d0099200ae1a455768d208 ]

Don't do extra cpu_to_le32 conversion for put_unaligned_le32 because it is already implemented in this function.

Fixes sparse error:
xhci-hub.c:1152:44: warning: incorrect type in argument 1 (different base types)
xhci-hub.c:1152:44: expected unsigned int [usertype] val
xhci-hub.c:1152:44: got restricted __le32 [usertype]

Fixes 395f540 "xhci: support new USB 3.1 hub request to get extended port status"
Cc Mathias Nyman <[email protected]>
---
 drivers/usb/host/xhci-hub.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/usb/host/xhci-hub.c +1 -1
index 8f180bf7561a..9772c0de59b7 100644
@@ -1104,7 +1104,7 @@ int xhci_hub_control(struct usb_hcd *hcd, u16 typeReq, u16 wValue,
 			}
 			port_li = readl(ports[wIndex]->addr + PORTLI);
 			status = xhci_get_ext_port_status(temp, port_li);
-			put_unaligned_le32(cpu_to_le32(status), &buf[4]);
+			put_unaligned_le32(status, &buf[4]);
 		}
 		break;
 	case SetPortFeature:
Signature · 1 line
2.20.1
author maintainer
#207 lore raw

PATCH 471/671 media: rcar-vin: Clean up correct notifier in error path

In reply to Sasha Levin

From: Niklas Söderlund <[email protected]>

[ Upstream commit 0bd465765f8d1300040de627f0a0971a8849d654 ]

The parallel input initialization error path cleans up the wrong async notifier, fix this by cleaning up the correct notifier.

---
 drivers/media/platform/rcar-vin/rcar-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/rcar-vin/rcar-core.c +1 -1
index 485fa3fa8b49..c389ba9ba74d 100644
@@ -631,7 +631,7 @@ static int rvin_parallel_init(struct rvin_dev *vin)
 	ret = v4l2_async_notifier_register(&vin->v4l2_dev, &vin->notifier);
 	if (ret < 0) {
 		vin_err(vin, "Notifier registration failed\n");
-		v4l2_async_notifier_cleanup(&vin->group->notifier);
+		v4l2_async_notifier_cleanup(&vin->notifier);
 		return ret;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#208 lore raw

PATCH 472/671 mic: avoid statically declaring a 'struct device'.

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit bc83f79bd2119230888fb8574639d5a51b38f903 ]

Generally, declaring a platform device as a static variable is a bad idea and can cause all kinds of problems, in particular with the DMA configuration and lifetime rules.

A specific problem we hit here is from a bug in clang that warns about certain (otherwise valid) macros when used in static variables:

drivers/misc/mic/card/mic_x100.c:285:27: warning: shift count >= width of type [-Wshift-count-overflow]

static u64 mic_dma_mask = DMA_BIT_MASK(64);
                          ^~~~~~~~~~~~~~~~
include/linux/dma-mapping.h:141:54: note: expanded from macro 'DMA_BIT_MASK'
 #define DMA_BIT_MASK(n) (((n) == 64) ? ~0ULL : ((1ULL<<(n))-1))
                                                     ^ ~~~

A slightly better way here is to create the platform device dynamically and set the dma mask in the probe function.
This avoids the warning and some other problems, but is still not ideal because the device creation should really be separated from the driver, and the fact that the device has no parent means we have to force the dma mask rather than having it set up from the bus that the device is actually on.

Fixes dd8d8d44df64 ("misc: mic: MIC card driver specific changes to enable SCIF")
---
 drivers/misc/mic/card/mic_x100.c | 28 ++++++++++++----------------
 1 file changed, 12 insertions(+), 16 deletions(-)
drivers/misc/mic/card/mic_x100.c +12 -16
index b9f0710ffa6b..4007adc666f3 100644
@@ -249,6 +249,9 @@ static int __init mic_probe(struct platform_device *pdev)
 	mdrv->dev = &pdev->dev;
 	snprintf(mdrv->name, sizeof(mic_driver_name), mic_driver_name);
 
+	/* FIXME: use dma_set_mask_and_coherent() and check result */
+	dma_coerce_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(64));
+
 	mdev->mmio.pa = MIC_X100_MMIO_BASE;
 	mdev->mmio.len = MIC_X100_MMIO_LEN;
 	mdev->mmio.va = devm_ioremap(&pdev->dev, MIC_X100_MMIO_BASE,
@@ -294,18 +297,6 @@ static void mic_platform_shutdown(struct platform_device *pdev)
 	mic_remove(pdev);
 }
 
-static u64 mic_dma_mask = DMA_BIT_MASK(64);
-
-static struct platform_device mic_platform_dev = {
-	.name = mic_driver_name,
-	.id   = 0,
-	.num_resources = 0,
-	.dev = {
-		.dma_mask = &mic_dma_mask,
-		.coherent_dma_mask = DMA_BIT_MASK(64),
-	},
-};
-
 static struct platform_driver __refdata mic_platform_driver = {
 	.probe = mic_probe,
 	.remove = mic_remove,
@@ -315,6 +306,8 @@ static struct platform_driver __refdata mic_platform_driver = {
 	},
 };
 
+static struct platform_device *mic_platform_dev;
+
 static int __init mic_init(void)
 {
 	int ret;
@@ -328,9 +321,12 @@ static int __init mic_init(void)
 
 	request_module("mic_x100_dma");
 	mic_init_card_debugfs();
-	ret = platform_device_register(&mic_platform_dev);
+
+	mic_platform_dev = platform_device_register_simple(mic_driver_name,
+							   0, NULL, 0);
+	ret = PTR_ERR_OR_ZERO(mic_platform_dev);
 	if (ret) {
-		pr_err("platform_device_register ret %d\n", ret);
+		pr_err("platform_device_register_full ret %d\n", ret);
 		goto cleanup_debugfs;
 	}
 	ret = platform_driver_register(&mic_platform_driver);
@@ -341,7 +337,7 @@ static int __init mic_init(void)
 	return ret;
 
 device_unregister:
-	platform_device_unregister(&mic_platform_dev);
+	platform_device_unregister(mic_platform_dev);
 cleanup_debugfs:
 	mic_exit_card_debugfs();
 done:
@@ -351,7 +347,7 @@ static int __init mic_init(void)
 static void __exit mic_exit(void)
 {
 	platform_driver_unregister(&mic_platform_driver);
-	platform_device_unregister(&mic_platform_dev);
+	platform_device_unregister(mic_platform_dev);
 	mic_exit_card_debugfs();
 }
 
Signature · 1 line
2.20.1
author maintainer
#209 lore raw

PATCH 473/671 x86/kgbd: Use NMI_VECTOR not APIC_DM_NMI

In reply to Sasha Levin

From: Thomas Gleixner <[email protected]>

[ Upstream commit 2591bc4e8d70b4e1330d327fb7e3921f4e070a51 ]

apic->send_IPI_allbutself() takes a vector number as argument.

APIC_DM_NMI is clearly not a vector number. It's defined to 0x400 which is outside the vector space.

Use NMI_VECTOR instead as that's what it is intended to be.

---
 arch/x86/kernel/kgdb.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/kgdb.c +1 -1
index 8e36f249646e..904e18bb38c5 100644
@@ -438,7 +438,7 @@ static void kgdb_disable_hw_debug(struct pt_regs *regs)
  */
 void kgdb_roundup_cpus(unsigned long flags)
 {
-	apic->send_IPI_allbutself(APIC_DM_NMI);
+	apic->send_IPI_allbutself(NMI_VECTOR);
 }
 #endif
 
Signature · 1 line
2.20.1
author maintainer
#210 lore raw

PATCH 474/671 crypto: ccp - Reduce maximum stack usage

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 72c8117adfced37df101c8c0b3f363e0906f83f0 ]

Each of the operations in ccp_run_cmd() needs several hundred bytes of kernel stack. Depending on the inlining, these may need separate stack slots that add up to more than the warning limit, as shown in this clang based build:

drivers/crypto/ccp/ccp-ops.c:871:12: error: stack frame size of 1164 bytes in function 'ccp_run_aes_cmd' [-Werror,-Wframe-larger-than=] static int ccp_run_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)

The problem may also happen when there is no warning, e.g. in the ccp_run_cmd()->ccp_run_aes_cmd()->ccp_run_aes_gcm_cmd() call chain with over 2000 bytes.

Mark each individual function as 'noinline_for_stack' to prevent this from happening, and move the calls to the two special cases for aes into the top-level function. This will keep the actual combined stack usage to the mimimum: 828 bytes for ccp_run_aes_gcm_cmd() and at most 524 bytes for each of the other cases.

Fixes 63b945091a07 ("crypto: ccp - CCP device driver and interface support")
---
 drivers/crypto/ccp/ccp-ops.c | 52 +++++++++++++++++++++---------------
 1 file changed, 31 insertions(+), 21 deletions(-)
drivers/crypto/ccp/ccp-ops.c +31 -21
index 4b48b8523a40..330853a2702f 100644
@@ -458,8 +458,8 @@ static int ccp_copy_from_sb(struct ccp_cmd_queue *cmd_q,
 	return ccp_copy_to_from_sb(cmd_q, wa, jobid, sb, byte_swap, true);
 }
 
-static int ccp_run_aes_cmac_cmd(struct ccp_cmd_queue *cmd_q,
-				struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_aes_cmac_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_aes_engine *aes = &cmd->u.aes;
 	struct ccp_dm_workarea key, ctx;
@@ -614,8 +614,8 @@ static int ccp_run_aes_cmac_cmd(struct ccp_cmd_queue *cmd_q,
 	return ret;
 }
 
-static int ccp_run_aes_gcm_cmd(struct ccp_cmd_queue *cmd_q,
-			       struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_aes_gcm_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_aes_engine *aes = &cmd->u.aes;
 	struct ccp_dm_workarea key, ctx, final_wa, tag;
@@ -897,7 +897,8 @@ static int ccp_run_aes_gcm_cmd(struct ccp_cmd_queue *cmd_q,
 	return ret;
 }
 
-static int ccp_run_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_aes_engine *aes = &cmd->u.aes;
 	struct ccp_dm_workarea key, ctx;
@@ -907,12 +908,6 @@ static int ccp_run_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 	bool in_place = false;
 	int ret;
 
-	if (aes->mode == CCP_AES_MODE_CMAC)
-		return ccp_run_aes_cmac_cmd(cmd_q, cmd);
-
-	if (aes->mode == CCP_AES_MODE_GCM)
-		return ccp_run_aes_gcm_cmd(cmd_q, cmd);
-
 	if (!((aes->key_len == AES_KEYSIZE_128) ||
 	      (aes->key_len == AES_KEYSIZE_192) ||
 	      (aes->key_len == AES_KEYSIZE_256)))
@@ -1080,8 +1075,8 @@ static int ccp_run_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 	return ret;
 }
 
-static int ccp_run_xts_aes_cmd(struct ccp_cmd_queue *cmd_q,
-			       struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_xts_aes_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_xts_aes_engine *xts = &cmd->u.xts;
 	struct ccp_dm_workarea key, ctx;
@@ -1280,7 +1275,8 @@ static int ccp_run_xts_aes_cmd(struct ccp_cmd_queue *cmd_q,
 	return ret;
 }
 
-static int ccp_run_des3_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_des3_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_des3_engine *des3 = &cmd->u.des3;
 
@@ -1476,7 +1472,8 @@ static int ccp_run_des3_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 	return ret;
 }
 
-static int ccp_run_sha_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_sha_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_sha_engine *sha = &cmd->u.sha;
 	struct ccp_dm_workarea ctx;
@@ -1820,7 +1817,8 @@ static int ccp_run_sha_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 	return ret;
 }
 
-static int ccp_run_rsa_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_rsa_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_rsa_engine *rsa = &cmd->u.rsa;
 	struct ccp_dm_workarea exp, src, dst;
@@ -1951,8 +1949,8 @@ static int ccp_run_rsa_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 	return ret;
 }
 
-static int ccp_run_passthru_cmd(struct ccp_cmd_queue *cmd_q,
-				struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_passthru_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_passthru_engine *pt = &cmd->u.passthru;
 	struct ccp_dm_workarea mask;
@@ -2083,7 +2081,8 @@ static int ccp_run_passthru_cmd(struct ccp_cmd_queue *cmd_q,
 	return ret;
 }
 
-static int ccp_run_passthru_nomap_cmd(struct ccp_cmd_queue *cmd_q,
+static noinline_for_stack int
+ccp_run_passthru_nomap_cmd(struct ccp_cmd_queue *cmd_q,
 				      struct ccp_cmd *cmd)
 {
 	struct ccp_passthru_nomap_engine *pt = &cmd->u.passthru_nomap;
@@ -2424,7 +2423,8 @@ static int ccp_run_ecc_pm_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 	return ret;
 }
 
-static int ccp_run_ecc_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
+static noinline_for_stack int
+ccp_run_ecc_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 {
 	struct ccp_ecc_engine *ecc = &cmd->u.ecc;
 
@@ -2461,7 +2461,17 @@ int ccp_run_cmd(struct ccp_cmd_queue *cmd_q, struct ccp_cmd *cmd)
 
 	switch (cmd->engine) {
 	case CCP_ENGINE_AES:
-		ret = ccp_run_aes_cmd(cmd_q, cmd);
+		switch (cmd->u.aes.mode) {
+		case CCP_AES_MODE_CMAC:
+			ret = ccp_run_aes_cmac_cmd(cmd_q, cmd);
+			break;
+		case CCP_AES_MODE_GCM:
+			ret = ccp_run_aes_gcm_cmd(cmd_q, cmd);
+			break;
+		default:
+			ret = ccp_run_aes_cmd(cmd_q, cmd);
+			break;
+		}
 		break;
 	case CCP_ENGINE_XTS_AES_128:
 		ret = ccp_run_xts_aes_cmd(cmd_q, cmd);
Signature · 1 line
2.20.1
author maintainer
#211 lore raw

PATCH 475/671 ALSA: aoa: onyx: always initialize register read value

In reply to Sasha Levin

From: Johannes Berg <[email protected]>

[ Upstream commit f474808acb3c4b30552d9c59b181244e0300d218 ]

A lot of places in the driver use onyx_read_register() without checking the return value, and it's been working OK for ~10 years or so, so probably never fails ... Rather than trying to check the return value everywhere, which would be relatively intrusive, at least make sure we don't use an uninitialized value.

Fixes f3d9478b2ce4 ("[ALSA] snd-aoa: add snd-aoa")
---
 sound/aoa/codecs/onyx.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
sound/aoa/codecs/onyx.c +3 -1
index d2d96ca082b7..6224fd3bbf7c 100644
@@ -74,8 +74,10 @@ static int onyx_read_register(struct onyx *onyx, u8 reg, u8 *value)
 		return 0;
 	}
 	v = i2c_smbus_read_byte_data(onyx->i2c, reg);
-	if (v < 0)
+	if (v < 0) {
+		*value = 0;
 		return -1;
+	}
 	*value = (u8)v;
 	onyx->cache[ONYX_REG_CONTROL-FIRSTREGISTER] = *value;
 	return 0;
Signature · 6 lines
2.20.1

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
author maintainer
#212 lore raw

PATCH 476/671 arm64: dts: renesas: r8a77995: Fix register range of display node

In reply to Sasha Levin

From: Yoshihiro Kaneko <[email protected]>

[ Upstream commit 56d651e890f3befd616b6962a862f5ffa1a514fa ]

Since the R8A77995 SoC uses DU{0,1}, the range from the base address to the 0x4000 address is used.
This patch fixed it.

Fixes 18f1a773e3f9e6d1 ("arm64: dts: renesas: r8a77995: add DU support")
---
 arch/arm64/boot/dts/renesas/r8a77995.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm64/boot/dts/renesas/r8a77995.dtsi +1 -1
index fe77bc43c447..fb3ecb2c385d 100644
@@ -938,7 +938,7 @@
 
 		du: display@feb00000 {
 			compatible = "renesas,du-r8a77995";
-			reg = <0 0xfeb00000 0 0x80000>;
+			reg = <0 0xfeb00000 0 0x40000>;
 			interrupts = <GIC_SPI 256 IRQ_TYPE_LEVEL_HIGH>,
 				     <GIC_SPI 268 IRQ_TYPE_LEVEL_HIGH>;
 			clocks = <&cpg CPG_MOD 724>,
Signature · 1 line
2.20.1
author maintainer
#213 lore raw

PATCH 477/671 tipc: reduce risk of wakeup queue starvation

In reply to Sasha Levin

From: Jon Maloy <[email protected]>

[ Upstream commit 7c5b42055964f587e55bd87ef334c3a27e95d144 ]

In commit 365ad353c256 ("tipc: reduce risk of user starvation during link congestion") we allowed senders to add exactly one list of extra buffers to the link backlog queues during link congestion (aka "oversubscription"). However, the criteria for when to stop adding wakeup messages to the input queue when the overload abates is inaccurate, and may cause starvation problems during very high load.

Currently, we stop adding wakeup messages after 10 total failed attempts where we find that there is no space left in the backlog queue for a certain importance level. The counter for this is accumulated across all levels, which may lead the algorithm to leave the loop prematurely, although there may still be plenty of space available at some levels. The result is sometimes that messages near the wakeup queue tail are not added to the input queue as they should be.

We now introduce a more exact algorithm, where we keep adding wakeup messages to a level as long as the backlog queue has free slots for the corresponding level, and stop at the moment there are no more such slots or when there are no more wakeup messages to dequeue.

Fixes 365ad35 ("tipc: reduce risk of user starvation during link congestion")
---
 net/tipc/link.c | 29 +++++++++++++++++++++--------
 1 file changed, 21 insertions(+), 8 deletions(-)
net/tipc/link.c +21 -8
index 0fbf8ea18ce0..cc9a0485536b 100644
@@ -830,18 +830,31 @@ static int link_schedule_user(struct tipc_link *l, struct tipc_msg *hdr)
  */
 static void link_prepare_wakeup(struct tipc_link *l)
 {
+	struct sk_buff_head *wakeupq = &l->wakeupq;
+	struct sk_buff_head *inputq = l->inputq;
 	struct sk_buff *skb, *tmp;
-	int imp, i = 0;
+	struct sk_buff_head tmpq;
+	int avail[5] = {0,};
+	int imp = 0;
+
+	__skb_queue_head_init(&tmpq);
 
-	skb_queue_walk_safe(&l->wakeupq, skb, tmp) {
+	for (; imp <= TIPC_SYSTEM_IMPORTANCE; imp++)
+		avail[imp] = l->backlog[imp].limit - l->backlog[imp].len;
+
+	skb_queue_walk_safe(wakeupq, skb, tmp) {
 		imp = TIPC_SKB_CB(skb)->chain_imp;
-		if (l->backlog[imp].len < l->backlog[imp].limit) {
-			skb_unlink(skb, &l->wakeupq);
-			skb_queue_tail(l->inputq, skb);
-		} else if (i++ > 10) {
-			break;
-		}
+		if (avail[imp] <= 0)
+			continue;
+		avail[imp]--;
+		__skb_unlink(skb, wakeupq);
+		__skb_queue_tail(&tmpq, skb);
 	}
+
+	spin_lock_bh(&inputq->lock);
+	skb_queue_splice_tail(&tmpq, inputq);
+	spin_unlock_bh(&inputq->lock);
+
 }
 
 void tipc_link_reset(struct tipc_link *l)
Signature · 1 line
2.20.1
author maintainer
#214 lore raw

PATCH 478/671 ARM: dts: stm32: add missing vdda-supply to adc on stm32h743i-eval

In reply to Sasha Levin

From: Fabrice Gasnier <[email protected]>

[ Upstream commit 493e84c5dc4d703d976b5875f5db22dae08a0782 ]

Add missing vdda-supply required by STM32 ADC.

Fixes 090992a9ca54 ("ARM: dts: stm32: enable ADC on stm32h743i-eval

board")

---
 arch/arm/boot/dts/stm32h743i-eval.dts | 1 +
 1 file changed, 1 insertion(+)
arch/arm/boot/dts/stm32h743i-eval.dts +1 -0
index 3f8e0c4a998d..5bf64e63cdf3 100644
@@ -79,6 +79,7 @@
 };
 
 &adc_12 {
+	vdda-supply = <&vdda>;
 	vref-supply = <&vdda>;
 	status = "okay";
 	adc1: adc@0 {
Signature · 1 line
2.20.1
author maintainer
#215 lore raw

PATCH 479/671 coredump: split pipe command whitespace before expanding template

In reply to Sasha Levin

From: Paul Wise <[email protected]>

[ Upstream commit 315c69261dd3fa12dbc830d4fa00d1fad98d3b03 ]

Save the offsets of the start of each argument to avoid having to update pointers to each argument after every corename krealloc and to avoid having to duplicate the memory for the dump command.

Executable names containing spaces were previously being expanded from %e or %E and then split in the middle of the filename. This is incorrect behaviour since an argument list can represent arguments with spaces.

The splitting could lead to extra arguments being passed to the core dump handler that it might have interpreted as options or ignored completely.

Core dump handlers that are not aware of this Linux kernel issue will be using %e or %E without considering that it may be split and so they will be vulnerable to processes with spaces in their names breaking their argument list. If their internals are otherwise well written, such as if they are written in shell but quote arguments, they will work better after this change than before. If they are not well written, then there is a slight chance of breakage depending on the details of the code but they will already be fairly broken by the split filenames.

Core dump handlers that are aware of this Linux kernel issue will be placing %e or %E as the last item in their core_pattern and then aggregating all of the remaining arguments into one, separated by spaces. Alternatively they will be obtaining the filename via other methods. Both of these will be compatible with the new arrangement.

A side effect from this change is that unknown template types (for example %z) result in an empty argument to the dump handler instead of the argument being dropped. This is a desired change as:

It is easier for dump handlers to process empty arguments than dropped ones, especially if they are written in shell or don't pass each template item with a preceding command-line option in order to differentiate between individual template types. Most core_patterns in the wild do not use options so they can confuse different template types (especially numeric ones) if an earlier one gets dropped in old kernels. If the kernel introduces a new template type and a core_pattern uses it, the core dump handler might not expect that the argument can be dropped in old kernels.

For example, this can result in security issues when %d is dropped in old kernels. This happened with the corekeeper package in Debian and resulted in the interface between corekeeper and Linux having to be rewritten to use command-line options to differentiate between template types.

The core_pattern for most core dump handlers is written by the handler author who would generally not insert unknown template types so this change should be compatible with all the core dump handlers that exist.

---
 fs/coredump.c | 44 +++++++++++++++++++++++++++++++++++++++-----
 1 file changed, 39 insertions(+), 5 deletions(-)
fs/coredump.c +39 -5
index 1e2c87acac9b..a49acb49dce0 100644
@@ -7,6 +7,7 @@
 #include <linux/stat.h>
 #include <linux/fcntl.h>
 #include <linux/swap.h>
+#include <linux/ctype.h>
 #include <linux/string.h>
 #include <linux/init.h>
 #include <linux/pagemap.h>
@@ -187,11 +188,13 @@ static int cn_print_exe_file(struct core_name *cn)
  * name into corename, which must have space for at least
  * CORENAME_MAX_SIZE bytes plus one byte for the zero terminator.
  */
-static int format_corename(struct core_name *cn, struct coredump_params *cprm)
+static int format_corename(struct core_name *cn, struct coredump_params *cprm,
+			   size_t **argv, int *argc)
 {
 	const struct cred *cred = current_cred();
 	const char *pat_ptr = core_pattern;
 	int ispipe = (*pat_ptr == '|');
+	bool was_space = false;
 	int pid_in_pattern = 0;
 	int err = 0;
 
@@ -201,12 +204,35 @@ static int format_corename(struct core_name *cn, struct coredump_params *cprm)
 		return -ENOMEM;
 	cn->corename[0] = '\0';
 
-	if (ispipe)
+	if (ispipe) {
+		int argvs = sizeof(core_pattern) / 2;
+		(*argv) = kmalloc_array(argvs, sizeof(**argv), GFP_KERNEL);
+		if (!(*argv))
+			return -ENOMEM;
+		(*argv)[(*argc)++] = 0;
 		++pat_ptr;
+	}
 
 	/* Repeat as long as we have more pattern to process and more output
 	   space */
 	while (*pat_ptr) {
+		/*
+		 * Split on spaces before doing template expansion so that
+		 * %e and %E don't get split if they have spaces in them
+		 */
+		if (ispipe) {
+			if (isspace(*pat_ptr)) {
+				was_space = true;
+				pat_ptr++;
+				continue;
+			} else if (was_space) {
+				was_space = false;
+				err = cn_printf(cn, "%c", '\0');
+				if (err)
+					return err;
+				(*argv)[(*argc)++] = cn->used;
+			}
+		}
 		if (*pat_ptr != '%') {
 			err = cn_printf(cn, "%c", *pat_ptr++);
 		} else {
@@ -546,6 +572,8 @@ void do_coredump(const siginfo_t *siginfo)
 	struct cred *cred;
 	int retval = 0;
 	int ispipe;
+	size_t *argv = NULL;
+	int argc = 0;
 	struct files_struct *displaced;
 	/* require nonrelative corefile path and be extra careful */
 	bool need_suid_safe = false;
@@ -592,9 +620,10 @@ void do_coredump(const siginfo_t *siginfo)
 
 	old_cred = override_creds(cred);
 
-	ispipe = format_corename(&cn, &cprm);
+	ispipe = format_corename(&cn, &cprm, &argv, &argc);
 
 	if (ispipe) {
+		int argi;
 		int dump_count;
 		char **helper_argv;
 		struct subprocess_info *sub_info;
@@ -637,12 +666,16 @@ void do_coredump(const siginfo_t *siginfo)
 			goto fail_dropcount;
 		}
 
-		helper_argv = argv_split(GFP_KERNEL, cn.corename, NULL);
+		helper_argv = kmalloc_array(argc + 1, sizeof(*helper_argv),
+					    GFP_KERNEL);
 		if (!helper_argv) {
 			printk(KERN_WARNING "%s failed to allocate memory\n",
 			       __func__);
 			goto fail_dropcount;
 		}
+		for (argi = 0; argi < argc; argi++)
+			helper_argv[argi] = cn.corename + argv[argi];
+		helper_argv[argi] = NULL;
 
 		retval = -ENOMEM;
 		sub_info = call_usermodehelper_setup(helper_argv[0],
@@ -652,7 +685,7 @@ void do_coredump(const siginfo_t *siginfo)
 			retval = call_usermodehelper_exec(sub_info,
 							  UMH_WAIT_EXEC);
 
-		argv_free(helper_argv);
+		kfree(helper_argv);
 		if (retval) {
 			printk(KERN_INFO "Core dump to |%s pipe failed\n",
 			       cn.corename);
@@ -766,6 +799,7 @@ void do_coredump(const siginfo_t *siginfo)
 	if (ispipe)
 		atomic_dec(&core_dump_count);
 fail_unlock:
+	kfree(argv);
 	kfree(cn.corename);
 	coredump_finish(mm, core_dumped);
 	revert_creds(old_cred);
Signature · 1 line
2.20.1
author maintainer
#216 lore raw

PATCH 480/671 net/mlx5: Fix mlx5_ifc_query_lag_out_bits

In reply to Sasha Levin

From: Mark Zhang <[email protected]>

[ Upstream commit ea77388b02270b0af8dc57f668f311235ea068f0 ]

Remove the "reserved_at_40" field to match the device specification.

Fixes 84df61ebc69b ("net/mlx5: Add HW interfaces used by LAG")
---
 include/linux/mlx5/mlx5_ifc.h | 2 --
 1 file changed, 2 deletions(-)
include/linux/mlx5/mlx5_ifc.h +0 -2
index 177f11c96187..76b76b6aa83d 100644
@@ -9053,8 +9053,6 @@ struct mlx5_ifc_query_lag_out_bits {
 
 	u8         syndrome[0x20];
 
-	u8         reserved_at_40[0x40];
-
 	struct mlx5_ifc_lagc_bits ctx;
 };
 
Signature · 1 line
2.20.1
author maintainer
#217 lore raw

PATCH 481/671 cifs: fix rmmod regression in cifs.ko caused by force_sig changes

In reply to Sasha Levin

From: Steve French <[email protected]>

[ Upstream commit 247bc9470b1eeefc7b58cdf2c39f2866ba651509 ]

Fixes 72abe3bcf091 ("signal/cifs: Fix cifs_put_tcp_session to call send_sig instead of force_sig")

The global change from force_sig caused module unloading of cifs.ko to fail (since the cifsd process could not be killed, "rmmod cifs" now would always fail)

---
 fs/cifs/connect.c | 1 +
 1 file changed, 1 insertion(+)
fs/cifs/connect.c +1 -0
index a59dcda07534..a8790bf04e95 100644
@@ -970,6 +970,7 @@ cifs_demultiplex_thread(void *p)
 		mempool_resize(cifs_req_poolp, length + cifs_min_rcv);
 
 	set_freezable();
+	allow_signal(SIGKILL);
 	while (server->tcpStatus != CifsExiting) {
 		if (try_to_freeze())
 			continue;
Signature · 1 line
2.20.1
author maintainer
#218 lore raw

PATCH 482/671 iio: tsl2772: Use devm_add_action_or_reset for tsl2772_chip_off

In reply to Sasha Levin

From: Chuhong Yuan <[email protected]>

[ Upstream commit 338084135aeddb103624a6841972fb8588295cc6 ]

Use devm_add_action_or_reset to call tsl2772_chip_off when the device is removed.
This also fixes the issue that the chip is turned off before the device is unregistered.

Not marked for stable as fairly hard to hit the bug and this is in the middle of a set making other cleanups to the driver. Hence will probably need explicit backporting.

Fixes c06c4d793584 ("staging: iio: tsl2x7x/tsl2772: move out of staging")
---
 drivers/iio/light/tsl2772.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)
drivers/iio/light/tsl2772.c +13 -3
index df5b2a0da96c..f2e308c6d6d7 100644
@@ -716,6 +716,13 @@ static int tsl2772_chip_off(struct iio_dev *indio_dev)
 	return tsl2772_write_control_reg(chip, 0x00);
 }
 
+static void tsl2772_chip_off_action(void *data)
+{
+	struct iio_dev *indio_dev = data;
+
+	tsl2772_chip_off(indio_dev);
+}
+
 /**
  * tsl2772_invoke_change - power cycle the device to implement the user
  *                         parameters
@@ -1711,9 +1718,14 @@ static int tsl2772_probe(struct i2c_client *clientp,
 	if (ret < 0)
 		return ret;
 
+	ret = devm_add_action_or_reset(&clientp->dev,
+					tsl2772_chip_off_action,
+					indio_dev);
+	if (ret < 0)
+		return ret;
+
 	ret = iio_device_register(indio_dev);
 	if (ret) {
-		tsl2772_chip_off(indio_dev);
 		dev_err(&clientp->dev,
 			"%s: iio registration failed\n", __func__);
 		return ret;
@@ -1740,8 +1752,6 @@ static int tsl2772_remove(struct i2c_client *client)
 {
 	struct iio_dev *indio_dev = i2c_get_clientdata(client);
 
-	tsl2772_chip_off(indio_dev);
-
 	iio_device_unregister(indio_dev);
 
 	return 0;
Signature · 1 line
2.20.1
Acked-by on v1 482/671
#219 lore raw

Re: [PATCH AUTOSEL 4.19 482/671] iio: tsl2772: Use devm_add_action_or_reset for tsl2772_chip_off

Quoted 4 lines · On Thu, 16 Jan 2020 12:02:00 -0500 Sasha Levin <[email protected]> wrote:

From: Chuhong Yuan <[email protected]>

[ Upstream commit 338084135aeddb103624a6841972fb8588295cc6 ]

Quoted 8 lines

Use devm_add_action_or_reset to call tsl2772_chip_off when the device is removed.
This also fixes the issue that the chip is turned off before the device is unregistered.

Not marked for stable as fairly hard to hit the bug and this is in the middle of a set making other cleanups to the driver. Hence will probably need explicit backporting.

Guess I was wrong and it does go on cleanly. I took a quick look at current 4.19 driver and looks like it's fine on it's own.

We need to be careful with this one in general though.

Quoted 54 lines
Fixes c06c4d793584 ("staging: iio: tsl2x7x/tsl2772: move out of staging")
---
 drivers/iio/light/tsl2772.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)
drivers/iio/light/tsl2772.c +13 -3
index df5b2a0da96c..f2e308c6d6d7 100644
@@ -716,6 +716,13 @@ static int tsl2772_chip_off(struct iio_dev *indio_dev)
 	return tsl2772_write_control_reg(chip, 0x00);
 }
 
+static void tsl2772_chip_off_action(void *data)
+{
+	struct iio_dev *indio_dev = data;
+
+	tsl2772_chip_off(indio_dev);
+}
+
 /**
  * tsl2772_invoke_change - power cycle the device to implement the user
  *                         parameters
@@ -1711,9 +1718,14 @@ static int tsl2772_probe(struct i2c_client *clientp,
 	if (ret < 0)
 		return ret;
 
+	ret = devm_add_action_or_reset(&clientp->dev,
+					tsl2772_chip_off_action,
+					indio_dev);
+	if (ret < 0)
+		return ret;
+
 	ret = iio_device_register(indio_dev);
 	if (ret) {
-		tsl2772_chip_off(indio_dev);
 		dev_err(&clientp->dev,
 			"%s: iio registration failed\n", __func__);
 		return ret;
@@ -1740,8 +1752,6 @@ static int tsl2772_remove(struct i2c_client *client)
 {
 	struct iio_dev *indio_dev = i2c_get_clientdata(client);
 
-	tsl2772_chip_off(indio_dev);
-
 	iio_device_unregister(indio_dev);
 
 	return 0;
author maintainer
#220 lore raw

Re: [PATCH AUTOSEL 4.19 482/671] iio: tsl2772: Use devm_add_action_or_reset for tsl2772_chip_off

Quoted 16 lines · Jonathan Cameron
Quoted 4 lines · On Thu, 16 Jan 2020 12:02:00 -0500 Sasha Levin <[email protected]> wrote:

From: Chuhong Yuan <[email protected]>

[ Upstream commit 338084135aeddb103624a6841972fb8588295cc6 ]

Quoted 8 lines

Use devm_add_action_or_reset to call tsl2772_chip_off when the device is removed.
This also fixes the issue that the chip is turned off before the device is unregistered.

Not marked for stable as fairly hard to hit the bug and this is in the middle of a set making other cleanups to the driver. Hence will probably need explicit backporting.

Quoted 7 lines

Guess I was wrong and it does go on cleanly. I took a quick look at current 4.19 driver and looks like it's fine on it's own.

We need to be careful with this one in general though.

Thanks Jonathan. I saw the comment, but it applied and built cleanly, and looked sane enough without any related changes.

Signature · 2 lines
Thanks,
Sasha
author maintainer
#221 lore raw

PATCH 483/671 net: fix bpf_xdp_adjust_head regression for generic-XDP

In reply to Sasha Levin

From: Jesper Dangaard Brouer <[email protected]>

[ Upstream commit 065af355470519bd184019a93ac579f22b036045 ]

When generic-XDP was moved to a later processing step by commit 458bf2f224f0 ("net: core: support XDP generic on stacked devices.") a regression was introduced when using bpf_xdp_adjust_head.

The issue is that after this commit the skb->network_header is now changed prior to calling generic XDP and not after. Thus, if the header is changed by XDP (via bpf_xdp_adjust_head), then skb->network_header also need to be updated again. Fix by calling skb_reset_network_header().

Fixes 458bf2f224f0 ("net: core: support XDP generic on stacked devices.")
---
 net/core/dev.c | 15 ++++++++++-----
 1 file changed, 10 insertions(+), 5 deletions(-)
net/core/dev.c +10 -5
index 935fe158cfaf..73ebacabfde8 100644
@@ -4349,12 +4349,17 @@ static u32 netif_receive_generic_xdp(struct sk_buff *skb,
 
 	act = bpf_prog_run_xdp(xdp_prog, xdp);
 
+	/* check if bpf_xdp_adjust_head was used */
 	off = xdp->data - orig_data;
-	if (off > 0)
-		__skb_pull(skb, off);
-	else if (off < 0)
-		__skb_push(skb, -off);
-	skb->mac_header += off;
+	if (off) {
+		if (off > 0)
+			__skb_pull(skb, off);
+		else if (off < 0)
+			__skb_push(skb, -off);
+
+		skb->mac_header += off;
+		skb_reset_network_header(skb);
+	}
 
 	/* check if bpf_xdp_adjust_tail was used. it can only "shrink"
 	 * pckt.
Signature · 1 line
2.20.1
author maintainer
#222 lore raw

PATCH 484/671 spi: bcm-qspi: Fix BSPI QUAD and DUAL mode support when using flex mode

In reply to Sasha Levin

From: Rayagonda Kokatanur <[email protected]>

[ Upstream commit 79629d0f7ce5b38515c1716911a0181f01b91102 ]

Fix data transfer width settings based on DT field 'spi-rx-bus-width' to configure BSPI in single, dual or quad mode by using data width and not the command width.

Fixes 5f195ee7d830c ("spi: bcm-qspi: Implement the spi_mem interface")
---
 drivers/spi/spi-bcm-qspi.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/spi/spi-bcm-qspi.c +2 -2
index 584bcb018a62..285a6f463013 100644
@@ -354,7 +354,7 @@ static int bcm_qspi_bspi_set_flex_mode(struct bcm_qspi *qspi,
 {
 	int bpc = 0, bpp = 0;
 	u8 command = op->cmd.opcode;
-	int width  = op->cmd.buswidth ? op->cmd.buswidth : SPI_NBITS_SINGLE;
+	int width = op->data.buswidth ? op->data.buswidth : SPI_NBITS_SINGLE;
 	int addrlen = op->addr.nbytes;
 	int flex_mode = 1;
 
@@ -992,7 +992,7 @@ static int bcm_qspi_exec_mem_op(struct spi_mem *mem,
 	if (mspi_read)
 		return bcm_qspi_mspi_exec_mem_op(spi, op);
 
-	ret = bcm_qspi_bspi_set_mode(qspi, op, -1);
+	ret = bcm_qspi_bspi_set_mode(qspi, op, 0);
 
 	if (!ret)
 		ret = bcm_qspi_bspi_exec_mem_op(spi, op);
Signature · 1 line
2.20.1
author maintainer
#223 lore raw

PATCH 485/671 cxgb4: smt: Add lock for atomic_dec_and_test

In reply to Sasha Levin

From: Chuhong Yuan <[email protected]>

[ Upstream commit 4a8937b83892cb69524291cae6cdabad4a8be033 ]

The atomic_dec_and_test() is not safe because it is outside of locks.
Move the locks of t4_smte_free() to its caller, cxgb4_smt_release() to protect the atomic decrement.

Fixes 3bdb376e6944 ("cxgb4: introduce SMT ops to prepare for SMAC rewrite support")
---
 drivers/net/ethernet/chelsio/cxgb4/smt.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/chelsio/cxgb4/smt.c +2 -2
index 7b2207a2a130..9b3f4205cb4d 100644
@@ -98,11 +98,9 @@ static struct smt_entry *find_or_alloc_smte(struct smt_data *s, u8 *smac)
 
 static void t4_smte_free(struct smt_entry *e)
 {
-	spin_lock_bh(&e->lock);
 	if (atomic_read(&e->refcnt) == 0) {  /* hasn't been recycled */
 		e->state = SMT_STATE_UNUSED;
 	}
-	spin_unlock_bh(&e->lock);
 }
 
 /**
@@ -112,8 +110,10 @@ static void t4_smte_free(struct smt_entry *e)
  */
 void cxgb4_smt_release(struct smt_entry *e)
 {
+	spin_lock_bh(&e->lock);
 	if (atomic_dec_and_test(&e->refcnt))
 		t4_smte_free(e);
+	spin_unlock_bh(&e->lock);
 }
 EXPORT_SYMBOL(cxgb4_smt_release);
 
Signature · 1 line
2.20.1
author maintainer
#224 lore raw

PATCH 486/671 crypto: caam - free resources in case caam_rng registration failed

In reply to Sasha Levin

From: Iuliana Prodan <[email protected]>

[ Upstream commit c59a1d41672a89b5cac49db1a472ff889e35a2d2 ]

Check the return value of the hardware registration for caam_rng and free resources in case of failure.

Fixes e24f7c9e87d4 ("crypto: caam - hwrng support")
---
 drivers/crypto/caam/caamrng.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/crypto/caam/caamrng.c +4 -1
index fde07d4ff019..ff6718a11e9e 100644
@@ -353,7 +353,10 @@ static int __init caam_rng_init(void)
 		goto free_rng_ctx;
 
 	dev_info(dev, "registering rng-caam\n");
-	return hwrng_register(&caam_rng);
+
+	err = hwrng_register(&caam_rng);
+	if (!err)
+		return err;
 
 free_rng_ctx:
 	kfree(rng_ctx);
Signature · 1 line
2.20.1
author maintainer
#225 lore raw

PATCH 487/671 ext4: set error return correctly when ext4_htree_store_dirent fails

In reply to Sasha Levin

From: Colin Ian King <[email protected]>

[ Upstream commit 7a14826ede1d714f0bb56de8167c0e519041eeda ]

Currently when the call to ext4_htree_store_dirent fails the error return variable 'ret' is is not being set to the error code and variable count is instead, hence the error code is not being returned. Fix this by assigning ret to the error return code.

Addresses-Coverity: ("Unused value")

Fixes 8af0f0822797 ("ext4: fix readdir error in the case of inline_data+dir_index")
---
 fs/ext4/inline.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/ext4/inline.c +1 -1
index 56f6e1782d5f..4572cb057951 100644
@@ -1419,7 +1419,7 @@ int htree_inlinedir_to_tree(struct file *dir_file,
 		err = ext4_htree_store_dirent(dir_file, hinfo->hash,
 					      hinfo->minor_hash, de, &tmp_str);
 		if (err) {
-			count = err;
+			ret = err;
 			goto out;
 		}
 		count++;
Signature · 1 line
2.20.1
author maintainer
#226 lore raw

PATCH 488/671 RDMA/hns: Bugfix for slab-out-of-bounds when unloading hip08 driver

In reply to Sasha Levin

From: Xi Wang <[email protected]>

[ Upstream commit 9bba3f0cbfc8abf2e1549ea03c0128186081d7a8 ]

kasan will report a BUG when run command 'rmmod hns_roce_hw_v2', the calltrace is as follows:

==================================================================

BUG: KASAN: slab-out-of-bounds in hns_roce_table_mhop_put+0x584/0x828
[hns_roce]
Read of size 8 at addr ffff802185e08300 by task rmmod/270

Call trace:
dump_backtrace+0x0/0x1e8
show_stack+0x14/0x20
dump_stack+0xc4/0xfc
print_address_description+0x60/0x270
__kasan_report+0x164/0x1b8
kasan_report+0xc/0x18
__asan_load8+0x84/0xa8
hns_roce_table_mhop_put+0x584/0x828 [hns_roce]
hns_roce_table_put+0x174/0x1a0 [hns_roce]
hns_roce_mr_free+0x124/0x210 [hns_roce]
hns_roce_dereg_mr+0x90/0xb8 [hns_roce]
ib_dealloc_pd_user+0x60/0xf0
ib_mad_port_close+0x128/0x1d8
ib_mad_remove_device+0x94/0x118
remove_client_context+0xa0/0xe0
disable_device+0xfc/0x1c0
__ib_unregister_device+0x60/0xe0
ib_unregister_device+0x24/0x38
hns_roce_exit+0x3c/0x138 [hns_roce]
__hns_roce_hw_v2_uninit_instance.isra.30+0x28/0x50 [hns_roce_hw_v2]
hns_roce_hw_v2_uninit_instance+0x44/0x60 [hns_roce_hw_v2]
hclge_uninit_client_instance+0x15c/0x238 [hclge]
hnae3_uninit_client_instance+0x84/0xa8 [hnae3]
hnae3_unregister_client+0x84/0x158 [hnae3]
hns_roce_hw_v2_exit+0x14/0x20 [hns_roce_hw_v2]
__arm64_sys_delete_module+0x20c/0x308
el0_svc_handler+0xbc/0x210
el0_svc+0x8/0xc

Allocated by task 255:
__kasan_kmalloc.isra.0+0xd0/0x180
kasan_kmalloc+0xc/0x18
__kmalloc+0x16c/0x328
hns_roce_init_hem_table+0x20c/0x428 [hns_roce]
hns_roce_init+0x214/0xfe0 [hns_roce]
__hns_roce_hw_v2_init_instance+0x284/0x330 [hns_roce_hw_v2]
hns_roce_hw_v2_init_instance+0xd0/0x1b8 [hns_roce_hw_v2]
hclge_init_roce_client_instance+0x180/0x310 [hclge]
hclge_init_client_instance+0xcc/0x508 [hclge]
hnae3_init_client_instance.part.3+0x3c/0x80 [hnae3]
hnae3_register_client+0x134/0x1a8 [hnae3]
0xffff200009c00014
do_one_initcall+0x9c/0x3e0
do_init_module+0xd4/0x2d8
load_module+0x3284/0x3690
__se_sys_init_module+0x274/0x308
__arm64_sys_init_module+0x40/0x50
el0_svc_handler+0xbc/0x210
el0_svc+0x8/0xc

Freed by task 0:

(stack is not available)

The buggy address belongs to the object at ffff802185e06300
which belongs to the cache kmalloc-8k of size 8192
The buggy address is located 0 bytes to the right of
8192-byte region [ffff802185e06300, ffff802185e08300)
The buggy address belongs to the page:
page:ffff7fe008617800 refcount:1 mapcount:0 mapping:ffff802340020e00 index:0x0
compound_mapcount: 0
flags: 0x5fffe00000010200(slab|head)
raw: 5fffe00000010200 dead000000000100 dead000000000200 ffff802340020e00
raw: 0000000000000000 00000000803e003e 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
ffff802185e08200: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
ffff802185e08280: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff802185e08300: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc

^
ffff802185e08380: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
ffff802185e08400: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
==================================================================
Disabling lock debugging due to kernel taint

Fixes a25d13cbe816 ("RDMA/hns: Add the interfaces to support multi hop addressing for the contexts in hip08")
---
 drivers/infiniband/hw/hns/hns_roce_hem.c | 19 +++++++++++++------
 1 file changed, 13 insertions(+), 6 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hem.c +13 -6
index a73d388b7093..31b9b99f81cb 100644
@@ -54,12 +54,13 @@ bool hns_roce_check_whether_mhop(struct hns_roce_dev *hr_dev, u32 type)
 EXPORT_SYMBOL_GPL(hns_roce_check_whether_mhop);
 
 static bool hns_roce_check_hem_null(struct hns_roce_hem **hem, u64 start_idx,
-			    u32 bt_chunk_num)
+			    u32 bt_chunk_num, u64 hem_max_num)
 {
-	int i;
+	u64 check_max_num = start_idx + bt_chunk_num;
+	u64 i;
 
-	for (i = 0; i < bt_chunk_num; i++)
-		if (hem[start_idx + i])
+	for (i = start_idx; (i < check_max_num) && (i < hem_max_num); i++)
+		if (hem[i])
 			return false;
 
 	return true;
@@ -413,6 +414,12 @@ static int hns_roce_table_mhop_get(struct hns_roce_dev *hr_dev,
 		return -EINVAL;
 	}
 
+	if (unlikely(hem_idx >= table->num_hem)) {
+		dev_err(dev, "Table %d exceed hem limt idx = %llu,max = %lu!\n",
+			     table->type, hem_idx, table->num_hem);
+		return -EINVAL;
+	}
+
 	mutex_lock(&table->mutex);
 
 	if (table->hem[hem_idx]) {
@@ -649,7 +656,7 @@ static void hns_roce_table_mhop_put(struct hns_roce_dev *hr_dev,
 	if (check_whether_bt_num_2(table->type, hop_num)) {
 		start_idx = mhop.l0_idx * chunk_ba_num;
 		if (hns_roce_check_hem_null(table->hem, start_idx,
-					    chunk_ba_num)) {
+					    chunk_ba_num, table->num_hem)) {
 			if (table->type < HEM_TYPE_MTT &&
 			    hr_dev->hw->clear_hem(hr_dev, table, obj, 0))
 				dev_warn(dev, "Clear HEM base address failed.\n");
@@ -663,7 +670,7 @@ static void hns_roce_table_mhop_put(struct hns_roce_dev *hr_dev,
 		start_idx = mhop.l0_idx * chunk_ba_num * chunk_ba_num +
 			    mhop.l1_idx * chunk_ba_num;
 		if (hns_roce_check_hem_null(table->hem, start_idx,
-					    chunk_ba_num)) {
+					    chunk_ba_num, table->num_hem)) {
 			if (hr_dev->hw->clear_hem(hr_dev, table, obj, 1))
 				dev_warn(dev, "Clear HEM base address failed.\n");
 
Signature · 1 line
2.20.1
author maintainer
#227 lore raw

PATCH 489/671 RDMA/hns: bugfix for slab-out-of-bounds when loading hip08 driver

In reply to Sasha Levin

From: Xi Wang <[email protected]>

[ Upstream commit bf8c02f961c89e5ccae5987b7ab28f5592a35101 ]

kasan will report a BUG when run command 'insmod hns_roce_hw_v2.ko', the calltrace is as follows:

==================================================================

BUG: KASAN: slab-out-of-bounds in hns_roce_v2_init_eq_table+0x1324/0x1948
[hns_roce_hw_v2]
Read of size 8 at addr ffff8020e7a10608 by task insmod/256

CPU: 0 PID: 256 Comm: insmod Tainted: G           O      5.2.0-rc4 #1
Hardware name: Huawei D06 /D06, BIOS Hisilicon D06 UEFI RC0
Call trace:
dump_backtrace+0x0/0x1e8
show_stack+0x14/0x20
dump_stack+0xc4/0xfc
print_address_description+0x60/0x270
__kasan_report+0x164/0x1b8
kasan_report+0xc/0x18
__asan_load8+0x84/0xa8
hns_roce_v2_init_eq_table+0x1324/0x1948 [hns_roce_hw_v2]
hns_roce_init+0xf8/0xfe0 [hns_roce]
__hns_roce_hw_v2_init_instance+0x284/0x330 [hns_roce_hw_v2]
hns_roce_hw_v2_init_instance+0xd0/0x1b8 [hns_roce_hw_v2]
hclge_init_roce_client_instance+0x180/0x310 [hclge]
hclge_init_client_instance+0xcc/0x508 [hclge]
hnae3_init_client_instance.part.3+0x3c/0x80 [hnae3]
hnae3_register_client+0x134/0x1a8 [hnae3]
hns_roce_hw_v2_init+0x14/0x10000 [hns_roce_hw_v2]
do_one_initcall+0x9c/0x3e0
do_init_module+0xd4/0x2d8
load_module+0x3284/0x3690
__se_sys_init_module+0x274/0x308
__arm64_sys_init_module+0x40/0x50
el0_svc_handler+0xbc/0x210
el0_svc+0x8/0xc

Allocated by task 256:
__kasan_kmalloc.isra.0+0xd0/0x180
kasan_kmalloc+0xc/0x18
__kmalloc+0x16c/0x328
hns_roce_v2_init_eq_table+0x764/0x1948 [hns_roce_hw_v2]
hns_roce_init+0xf8/0xfe0 [hns_roce]
__hns_roce_hw_v2_init_instance+0x284/0x330 [hns_roce_hw_v2]
hns_roce_hw_v2_init_instance+0xd0/0x1b8 [hns_roce_hw_v2]
hclge_init_roce_client_instance+0x180/0x310 [hclge]
hclge_init_client_instance+0xcc/0x508 [hclge]
hnae3_init_client_instance.part.3+0x3c/0x80 [hnae3]
hnae3_register_client+0x134/0x1a8 [hnae3]
hns_roce_hw_v2_init+0x14/0x10000 [hns_roce_hw_v2]
do_one_initcall+0x9c/0x3e0
do_init_module+0xd4/0x2d8
load_module+0x3284/0x3690
__se_sys_init_module+0x274/0x308
__arm64_sys_init_module+0x40/0x50
el0_svc_handler+0xbc/0x210
el0_svc+0x8/0xc

Freed by task 0:

(stack is not available)

The buggy address belongs to the object at ffff8020e7a10600
which belongs to the cache kmalloc-128 of size 128
The buggy address is located 8 bytes inside of
128-byte region [ffff8020e7a10600, ffff8020e7a10680)
The buggy address belongs to the page:
page:ffff7fe00839e840 refcount:1 mapcount:0 mapping:ffff802340020200 index:0x0
flags: 0x5fffe00000000200(slab)
raw: 5fffe00000000200 dead000000000100 dead000000000200 ffff802340020200
raw: 0000000000000000 0000000081000100 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
ffff8020e7a10500: 00 00 00 00 00 00 00 00 fc fc fc fc fc fc fc fc
ffff8020e7a10580: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
>ffff8020e7a10600: 00 fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc

^
ffff8020e7a10680: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
ffff8020e7a10700: 00 fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
==================================================================
Disabling lock debugging due to kernel taint

Fixes a5073d6054f7 ("RDMA/hns: Add eq support of hip08")
---
 drivers/infiniband/hw/hns/hns_roce_hw_v2.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/infiniband/hw/hns/hns_roce_hw_v2.c +4 -2
index 7021444f18b4..417de7ac0d5e 100644
@@ -4833,7 +4833,8 @@ static int hns_roce_mhop_alloc_eq(struct hns_roce_dev *hr_dev,
 				break;
 		}
 		eq->cur_eqe_ba = eq->buf_dma[0];
-		eq->nxt_eqe_ba = eq->buf_dma[1];
+		if (ba_num > 1)
+			eq->nxt_eqe_ba = eq->buf_dma[1];
 
 	} else if (mhop_num == 2) {
 		/* alloc L1 BT and buf */
@@ -4875,7 +4876,8 @@ static int hns_roce_mhop_alloc_eq(struct hns_roce_dev *hr_dev,
 				break;
 		}
 		eq->cur_eqe_ba = eq->buf_dma[0];
-		eq->nxt_eqe_ba = eq->buf_dma[1];
+		if (ba_num > 1)
+			eq->nxt_eqe_ba = eq->buf_dma[1];
 	}
 
 	eq->l0_last_num = i + 1;
Signature · 1 line
2.20.1
author maintainer
#228 lore raw

PATCH 490/671 ASoC: es8328: Fix copy-paste error in es8328_right_line_controls

In reply to Sasha Levin

From: YueHaibing <[email protected]>

[ Upstream commit 630742c296341a8cfe00dfd941392025ba8dd4e8 ]

It seems 'es8328_rline_enum' should be used
in es8328_right_line_controls

Fixes 567e4f98922c ("ASoC: add es8328 codec driver")
---
 sound/soc/codecs/es8328.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
sound/soc/codecs/es8328.c +1 -1
index e9fc2fd97d2f..3afa163f7652 100644
@@ -231,7 +231,7 @@ static const struct soc_enum es8328_rline_enum =
 			      ARRAY_SIZE(es8328_line_texts),
 			      es8328_line_texts);
 static const struct snd_kcontrol_new es8328_right_line_controls =
-	SOC_DAPM_ENUM("Route", es8328_lline_enum);
+	SOC_DAPM_ENUM("Route", es8328_rline_enum);
 
 /* Left Mixer */
 static const struct snd_kcontrol_new es8328_left_mixer_controls[] = {
Signature · 6 lines
2.20.1

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
author maintainer
#229 lore raw

PATCH 491/671 ASoC: cs4349: Use PM ops 'cs4349_runtime_pm'

In reply to Sasha Levin

From: YueHaibing <[email protected]>

[ Upstream commit 9b4275c415acca6264a3d7f1182589959c93d530 ]

sound/soc/codecs/cs4349.c:358:32: warning:
cs4349_runtime_pm defined but not used [-Wunused-const-variable=]

cs4349_runtime_pm ops already defined, it seems we should enable it.

Fixes e40da86 ("ASoC: cs4349: Add support for Cirrus Logic CS4349")
---
 sound/soc/codecs/cs4349.c | 1 +
 1 file changed, 1 insertion(+)
sound/soc/codecs/cs4349.c +1 -0
index bee0e343723f..f08d7a296c0c 100644
@@ -381,6 +381,7 @@ static struct i2c_driver cs4349_i2c_driver = {
 	.driver = {
 		.name		= "cs4349",
 		.of_match_table	= cs4349_of_match,
+		.pm = &cs4349_runtime_pm,
 	},
 	.id_table	= cs4349_i2c_id,
 	.probe		= cs4349_i2c_probe,
Signature · 1 line
2.20.1
author maintainer
#230 lore raw

PATCH 492/671 ASoC: wm8737: Fix copy-paste error in wm8737_snd_controls

In reply to Sasha Levin

From: YueHaibing <[email protected]>

[ Upstream commit 554b75bde64bcad9662530726d1483f7ef012069 ]

sound/soc/codecs/wm8737.c:112:29: warning:
high_3d defined but not used [-Wunused-const-variable=]

'high_3d' should be used for 3D High Cut-off.

---
 sound/soc/codecs/wm8737.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
sound/soc/codecs/wm8737.c +1 -1
index e9ae821e7609..568b77692f5f 100644
@@ -170,7 +170,7 @@ SOC_DOUBLE("Polarity Invert Switch", WM8737_ADC_CONTROL, 5, 6, 1, 0),
 SOC_SINGLE("3D Switch", WM8737_3D_ENHANCE, 0, 1, 0),
 SOC_SINGLE("3D Depth", WM8737_3D_ENHANCE, 1, 15, 0),
 SOC_ENUM("3D Low Cut-off", low_3d),
-SOC_ENUM("3D High Cut-off", low_3d),
+SOC_ENUM("3D High Cut-off", high_3d),
 SOC_SINGLE_TLV("3D ADC Volume", WM8737_3D_ENHANCE, 7, 1, 1, adc_tlv),
 
 SOC_SINGLE("Noise Gate Switch", WM8737_NOISE_GATE, 0, 1, 0),
Signature · 6 lines
2.20.1

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
author maintainer
#231 lore raw

PATCH 493/671 net/rds: Add a few missing rds_stat_names entries

In reply to Sasha Levin

From: Gerd Rausch <[email protected]>

[ Upstream commit 55c70ca00c982fbc0df4c4d3e31747fb73f4ddb5 ]

In a previous commit, fields were added to "struct rds_statistics" but array "rds_stat_names" was not updated accordingly.

Please note the inconsistent naming of the string representations that is done in the name of compatibility
with the Oracle internal code-base.

s_recv_bytes_added_to_socket -> "recv_bytes_added_to_sock"
s_recv_bytes_removed_from_socket -> "recv_bytes_freed_fromsock"

Fixes 192a798f5299 ("RDS: add stat for socket recv memory usage")
---
 net/rds/stats.c | 2 ++
 1 file changed, 2 insertions(+)
net/rds/stats.c +2 -0
index 73be187d389e..6bbab4d74c4f 100644
@@ -76,6 +76,8 @@ static const char *const rds_stat_names[] = {
 	"cong_update_received",
 	"cong_send_error",
 	"cong_send_blocked",
+	"recv_bytes_added_to_sock",
+	"recv_bytes_freed_fromsock",
 };
 
 void rds_stats_info_copy(struct rds_info_iterator *iter,
Signature · 1 line
2.20.1
author maintainer
#232 lore raw

PATCH 494/671 tools: bpftool: fix arguments for p_err() in do_event_pipe()

In reply to Sasha Levin

From: Quentin Monnet <[email protected]>

[ Upstream commit 9def249dc8409ffc1f5a1d7195f1c462f2b49c07 ]

The last argument passed to some calls to the p_err() functions is not correct, it should be "*argv" instead of "**argv". This may lead to a segmentation fault error if CPU IDs or indices from the command line cannot be parsed correctly. Let's fix this.

Fixes f412eed9dfde ("tools: bpftool: add simple perf event output reader")
---
 tools/bpf/bpftool/map_perf_ring.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
tools/bpf/bpftool/map_perf_ring.c +2 -2
index 6d41323be291..8ec0148d7426 100644
@@ -205,7 +205,7 @@ int do_event_pipe(int argc, char **argv)
 			NEXT_ARG();
 			cpu = strtoul(*argv, &endptr, 0);
 			if (*endptr) {
-				p_err("can't parse %s as CPU ID", **argv);
+				p_err("can't parse %s as CPU ID", *argv);
 				goto err_close_map;
 			}
 
@@ -216,7 +216,7 @@ int do_event_pipe(int argc, char **argv)
 			NEXT_ARG();
 			index = strtoul(*argv, &endptr, 0);
 			if (*endptr) {
-				p_err("can't parse %s as index", **argv);
+				p_err("can't parse %s as index", *argv);
 				goto err_close_map;
 			}
 
Signature · 1 line
2.20.1
author maintainer
#233 lore raw

PATCH 495/671 tools: bpftool: fix format strings and arguments for jsonw_printf()

In reply to Sasha Levin

From: Quentin Monnet <[email protected]>

[ Upstream commit 22c349e8db89df86804d3ba23cef037ccd44a8bf ]

There are some mismatches between format strings and arguments passed to jsonw_printf() in the BTF dumper for bpftool, which seems harmless but may result in warnings if the "__printf()" attribute is used correctly for jsonw_printf(). Let's fix relevant format strings and type cast.

Fixes b12d6ec09730 ("bpf: btf: add btf print functionality")
---
 tools/bpf/bpftool/btf_dumper.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
tools/bpf/bpftool/btf_dumper.c +4 -4
index e4e6e2b3fd84..ff0cc3c17141 100644
@@ -26,9 +26,9 @@ static void btf_dumper_ptr(const void *data, json_writer_t *jw,
 			   bool is_plain_text)
 {
 	if (is_plain_text)
-		jsonw_printf(jw, "%p", *(unsigned long *)data);
+		jsonw_printf(jw, "%p", data);
 	else
-		jsonw_printf(jw, "%u", *(unsigned long *)data);
+		jsonw_printf(jw, "%lu", *(unsigned long *)data);
 }
 
 static int btf_dumper_modifier(const struct btf_dumper *d, __u32 type_id,
@@ -129,7 +129,7 @@ static int btf_dumper_int(const struct btf_type *t, __u8 bit_offset,
 	switch (BTF_INT_ENCODING(*int_type)) {
 	case 0:
 		if (BTF_INT_BITS(*int_type) == 64)
-			jsonw_printf(jw, "%lu", *(__u64 *)data);
+			jsonw_printf(jw, "%llu", *(__u64 *)data);
 		else if (BTF_INT_BITS(*int_type) == 32)
 			jsonw_printf(jw, "%u", *(__u32 *)data);
 		else if (BTF_INT_BITS(*int_type) == 16)
@@ -142,7 +142,7 @@ static int btf_dumper_int(const struct btf_type *t, __u8 bit_offset,
 		break;
 	case BTF_INT_SIGNED:
 		if (BTF_INT_BITS(*int_type) == 64)
-			jsonw_printf(jw, "%ld", *(long long *)data);
+			jsonw_printf(jw, "%lld", *(long long *)data);
 		else if (BTF_INT_BITS(*int_type) == 32)
 			jsonw_printf(jw, "%d", *(int *)data);
 		else if (BTF_INT_BITS(*int_type) == 16)
Signature · 1 line
2.20.1
author maintainer
#234 lore raw

PATCH 496/671 drm: rcar-du: lvds: Fix bridge_to_rcar_lvds

In reply to Sasha Levin

From: Fabrizio Castro <[email protected]>

[ Upstream commit 0b936e6122738f4cf474d1f3ff636cba0edc8b94 ]

Using name "bridge" for macro bridge_to_rcar_lvds argument doesn't work when the pointer name used by the caller is not "bridge". Rename the argument to "b" to allow for any pointer name.

While at it, fix the connector_to_rcar_lvds macro similarly.

Fixes c6a27fa41fab ("drm: rcar-du: Convert LVDS encoder code to bridge driver")
[Fix connector_to_rcar_lvds]
---
 drivers/gpu/drm/rcar-du/rcar_lvds.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/gpu/drm/rcar-du/rcar_lvds.c +4 -4
index 4c39de3f4f0f..b6dc91cdff68 100644
@@ -59,11 +59,11 @@ struct rcar_lvds {
 	enum rcar_lvds_mode mode;
 };
 
-#define bridge_to_rcar_lvds(bridge) \
-	container_of(bridge, struct rcar_lvds, bridge)
+#define bridge_to_rcar_lvds(b) \
+	container_of(b, struct rcar_lvds, bridge)
 
-#define connector_to_rcar_lvds(connector) \
-	container_of(connector, struct rcar_lvds, connector)
+#define connector_to_rcar_lvds(c) \
+	container_of(c, struct rcar_lvds, connector)
 
 static void rcar_lvds_write(struct rcar_lvds *lvds, u32 reg, u32 data)
 {
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author maintainer
#235 lore raw

PATCH 497/671 bnxt_en: Fix handling FRAG_ERR when NVM_INSTALL_UPDATE cmd fails

In reply to Sasha Levin

From: Vasundhara Volam <[email protected]>

[ Upstream commit dd2ebf3404c7c295014bc025dea23960960ceb1a ]

If FW returns FRAG_ERR in response error code, driver is resending the command only when HWRM command returns success. Fix the code to resend NVM_INSTALL_UPDATE command with DEFRAG install flags, if FW returns FRAG_ERR in its response error code.

Fixes cb4d1d626145 ("bnxt_en: Retry failed NVM_INSTALL_UPDATE with defragmentation flag enabled.")
---
 drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)
drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +5 -7
index dc63d269f01d..cdbb8940a4ae 100644
@@ -1778,21 +1778,19 @@ static int bnxt_flash_package_from_file(struct net_device *dev,
 	mutex_lock(&bp->hwrm_cmd_lock);
 	hwrm_err = _hwrm_send_message(bp, &install, sizeof(install),
 				      INSTALL_PACKAGE_TIMEOUT);
-	if (hwrm_err)
-		goto flash_pkg_exit;
-
-	if (resp->error_code) {
+	if (hwrm_err) {
 		u8 error_code = ((struct hwrm_err_output *)resp)->cmd_err;
 
-		if (error_code == NVM_INSTALL_UPDATE_CMD_ERR_CODE_FRAG_ERR) {
+		if (resp->error_code && error_code ==
+		    NVM_INSTALL_UPDATE_CMD_ERR_CODE_FRAG_ERR) {
 			install.flags |= cpu_to_le16(
 			       NVM_INSTALL_UPDATE_REQ_FLAGS_ALLOWED_TO_DEFRAG);
 			hwrm_err = _hwrm_send_message(bp, &install,
 						      sizeof(install),
 						      INSTALL_PACKAGE_TIMEOUT);
-			if (hwrm_err)
-				goto flash_pkg_exit;
 		}
+		if (hwrm_err)
+			goto flash_pkg_exit;
 	}
 
 	if (resp->result) {
Signature · 1 line
2.20.1
author maintainer
#236 lore raw

PATCH 498/671 signal: Allow cifs and drbd to receive their terminating signals

In reply to Sasha Levin

From: "Eric W. Biederman" <[email protected]>

[ Upstream commit 33da8e7c814f77310250bb54a9db36a44c5de784 ]

My recent to change to only use force_sig for a synchronous events wound up breaking signal reception cifs and drbd. I had overlooked the fact that by default kthreads start out with all signals set to SIG_IGN. So a change I thought was safe turned out to have made it impossible for those kernel thread to catch their signals.

Reverting the work on force_sig is a bad idea because what the code was doing was very much a misuse of force_sig. As the way force_sig ultimately allowed the signal to happen was to change the signal handler to SIG_DFL. Which after the first signal will allow userspace to send signals to these kernel threads. At least for wake_ack_receiver in drbd that does not appear actively wrong.

So correct this problem by adding allow_kernel_signal that will allow signals whose siginfo reports they were sent by the kernel through, but will not allow userspace generated signals, and update cifs and drbd to call allow_kernel_signal in an appropriate place so that their thread can receive this signal.

Fixing things this way ensures that userspace won't be able to send signals and cause problems, that it is clear which signals the threads are expecting to receive, and it guarantees that nothing else in the system will be affected.

This change was partly inspired by similar cifs and drbd patches that added allow_signal.

Cc Steve French <[email protected]>
Cc Philipp Reisner <[email protected]>
Cc David Laight <[email protected]>
Fixes 247bc9470b1e ("cifs: fix rmmod regression in cifs.ko caused by force_sig changes")
Fixes 72abe3bcf091 ("signal/cifs: Fix cifs_put_tcp_session to call send_sig instead of force_sig")
Fixes fee109901f39 ("signal/drbd: Use send_sig not force_sig")
Fixes 3cf5d076fb4d ("signal: Remove task parameter from force_sig")
---
 drivers/block/drbd/drbd_main.c |  2 ++
 fs/cifs/connect.c              |  2 +-
 include/linux/signal.h         | 15 ++++++++++++++-
 kernel/signal.c                |  5 +++++
 4 files changed, 22 insertions(+), 2 deletions(-)
drivers/block/drbd/drbd_main.c +2 -0
index a49a8d91a599..5e3885f5729b 100644
@@ -334,6 +334,8 @@ static int drbd_thread_setup(void *arg)
 		 thi->name[0],
 		 resource->name);
 
+	allow_kernel_signal(DRBD_SIGKILL);
+	allow_kernel_signal(SIGXCPU);
 restart:
 	retval = thi->function(thi);
 
fs/cifs/connect.c +1 -1
index a8790bf04e95..576cf71576da 100644
@@ -970,7 +970,7 @@ cifs_demultiplex_thread(void *p)
 		mempool_resize(cifs_req_poolp, length + cifs_min_rcv);
 
 	set_freezable();
-	allow_signal(SIGKILL);
+	allow_kernel_signal(SIGKILL);
 	while (server->tcpStatus != CifsExiting) {
 		if (try_to_freeze())
 			continue;
include/linux/signal.h +14 -1
index e4d01469ed60..0be5ce2375cb 100644
@@ -272,6 +272,9 @@ extern void signal_setup_done(int failed, struct ksignal *ksig, int stepping);
 extern void exit_signals(struct task_struct *tsk);
 extern void kernel_sigaction(int, __sighandler_t);
 
+#define SIG_KTHREAD ((__force __sighandler_t)2)
+#define SIG_KTHREAD_KERNEL ((__force __sighandler_t)3)
+
 static inline void allow_signal(int sig)
 {
 	/*
@@ -279,7 +282,17 @@ static inline void allow_signal(int sig)
 	 * know it'll be handled, so that they don't get converted to
 	 * SIGKILL or just silently dropped.
 	 */
-	kernel_sigaction(sig, (__force __sighandler_t)2);
+	kernel_sigaction(sig, SIG_KTHREAD);
+}
+
+static inline void allow_kernel_signal(int sig)
+{
+	/*
+	 * Kernel threads handle their own signals. Let the signal code
+	 * know signals sent by the kernel will be handled, so that they
+	 * don't get silently dropped.
+	 */
+	kernel_sigaction(sig, SIG_KTHREAD_KERNEL);
 }
 
 static inline void disallow_signal(int sig)
kernel/signal.c +5 -0
index 7278302e3485..08911bb6fe9a 100644
@@ -86,6 +86,11 @@ static bool sig_task_ignored(struct task_struct *t, int sig, bool force)
 	    handler == SIG_DFL && !(force && sig_kernel_only(sig)))
 		return true;
 
+	/* Only allow kernel generated signals to this kthread */
+	if (unlikely((t->flags & PF_KTHREAD) &&
+		     (handler == SIG_KTHREAD_KERNEL) && !force))
+		return true;
+
 	return sig_handler_ignored(handler, sig);
 }
 
Signature · 1 line
2.20.1
author maintainer
#237 lore raw

PATCH 499/671 powerpc/64s/radix: Fix memory hot-unplug page table split

In reply to Sasha Levin

From: Nicholas Piggin <[email protected]>

[ Upstream commit 31f210cf42d4b308eacef89b6cb0b1459338b8de ]

create_physical_mapping expects physical addresses, but splitting these mapping on hot unplug is supplying virtual (effective) addresses.

Fixes 4dd5f8a99e791 ("powerpc/mm/radix: Split linear mapping on hot-unplug")
---
 arch/powerpc/mm/pgtable-radix.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/powerpc/mm/pgtable-radix.c +2 -2
index 69caeb5bccb2..5404a631d583 100644
@@ -717,8 +717,8 @@ static int __meminit stop_machine_change_mapping(void *data)
 
 	spin_unlock(&init_mm.page_table_lock);
 	pte_clear(&init_mm, params->aligned_start, params->pte);
-	create_physical_mapping(params->aligned_start, params->start, -1);
-	create_physical_mapping(params->end, params->aligned_end, -1);
+	create_physical_mapping(__pa(params->aligned_start), __pa(params->start), -1);
+	create_physical_mapping(__pa(params->end), __pa(params->aligned_end), -1);
 	spin_lock(&init_mm.page_table_lock);
 	return 0;
 }
Signature · 1 line
2.20.1
author maintainer
#238 lore raw

PATCH 500/671 ASoC: sun4i-i2s: RX and TX counter registers are swapped

In reply to Sasha Levin

From: Maxime Ripard <[email protected]>

[ Upstream commit cf2c0e1ce9544df42170fb921f12da82dc0cc8d6 ]

The RX and TX counters registers offset have been swapped, fix that.

---
 sound/soc/sunxi/sun4i-i2s.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
sound/soc/sunxi/sun4i-i2s.c +2 -2
index 18cf8404d27c..f248e563986c 100644
@@ -80,8 +80,8 @@
 #define SUN4I_I2S_CLK_DIV_MCLK_MASK		GENMASK(3, 0)
 #define SUN4I_I2S_CLK_DIV_MCLK(mclk)			((mclk) << 0)
 
-#define SUN4I_I2S_RX_CNT_REG		0x28
-#define SUN4I_I2S_TX_CNT_REG		0x2c
+#define SUN4I_I2S_TX_CNT_REG		0x28
+#define SUN4I_I2S_RX_CNT_REG		0x2c
 
 #define SUN4I_I2S_TX_CHAN_SEL_REG	0x30
 #define SUN4I_I2S_CHAN_SEL(num_chan)		(((num_chan) - 1) << 0)
Signature · 6 lines
2.20.1

_______________________________________________
Alsa-devel mailing list
[email protected]
https://mailman.alsa-project.org/mailman/listinfo/alsa-devel
author maintainer
#239 lore raw

PATCH 501/671 dmaengine: dw: platform: Switch to acpi_dma_controller_register()

In reply to Sasha Levin

From: Andy Shevchenko <[email protected]>

[ Upstream commit e7b8514e4d68bec21fc6385fa0a66797ddc34ac9 ]

There is a possibility to have registered ACPI DMA controller while it has been gone already.

To avoid the potential crash, move to non-managed acpi_dma_controller_register().

---
 drivers/dma/dw/platform.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)
drivers/dma/dw/platform.c +12 -2
index c299ff181bb6..62218ea0894c 100644
@@ -87,13 +87,20 @@ static void dw_dma_acpi_controller_register(struct dw_dma *dw)
 	dma_cap_set(DMA_SLAVE, info->dma_cap);
 	info->filter_fn = dw_dma_acpi_filter;
 
-	ret = devm_acpi_dma_controller_register(dev, acpi_dma_simple_xlate,
-						info);
+	ret = acpi_dma_controller_register(dev, acpi_dma_simple_xlate, info);
 	if (ret)
 		dev_err(dev, "could not register acpi_dma_controller\n");
 }
+
+static void dw_dma_acpi_controller_free(struct dw_dma *dw)
+{
+	struct device *dev = dw->dma.dev;
+
+	acpi_dma_controller_free(dev);
+}
 #else /* !CONFIG_ACPI */
 static inline void dw_dma_acpi_controller_register(struct dw_dma *dw) {}
+static inline void dw_dma_acpi_controller_free(struct dw_dma *dw) {}
 #endif /* !CONFIG_ACPI */
 
 #ifdef CONFIG_OF
@@ -249,6 +256,9 @@ static int dw_remove(struct platform_device *pdev)
 {
 	struct dw_dma_chip *chip = platform_get_drvdata(pdev);
 
+	if (ACPI_HANDLE(&pdev->dev))
+		dw_dma_acpi_controller_free(chip->dw);
+
 	if (pdev->dev.of_node)
 		of_dma_controller_free(pdev->dev.of_node);
 
Signature · 1 line
2.20.1
author maintainer
#240 lore raw

PATCH 502/671 rtc: rv3029: revert error handling patch to rv3029_eeprom_write()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit a6f26606ddd03c5eab8b2132f1bfaa768c06158f ]

My error handling "cleanup" was totally wrong. Both the "err" and "ret" variables are required. The "err" variable holds the error codes for rv3029_eeprom_enter/exit() and the "ret" variable holds the error codes for if actual write fails. In my patch if the write failed, the function probably still returned success.

Fixes 97f5b0379c38 ("rtc: rv3029: Clean up error handling in rv3029_eeprom_write()")
---
 drivers/rtc/rtc-rv3029c2.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/rtc/rtc-rv3029c2.c +8 -8
index 3d6174eb32f6..cfe3aece51d1 100644
@@ -282,13 +282,13 @@ static int rv3029_eeprom_read(struct device *dev, u8 reg,
 static int rv3029_eeprom_write(struct device *dev, u8 reg,
 			       u8 const buf[], size_t len)
 {
-	int ret;
+	int ret, err;
 	size_t i;
 	u8 tmp;
 
-	ret = rv3029_eeprom_enter(dev);
-	if (ret < 0)
-		return ret;
+	err = rv3029_eeprom_enter(dev);
+	if (err < 0)
+		return err;
 
 	for (i = 0; i < len; i++, reg++) {
 		ret = rv3029_read_regs(dev, reg, &tmp, 1);
@@ -304,11 +304,11 @@ static int rv3029_eeprom_write(struct device *dev, u8 reg,
 			break;
 	}
 
-	ret = rv3029_eeprom_exit(dev);
-	if (ret < 0)
-		return ret;
+	err = rv3029_eeprom_exit(dev);
+	if (err < 0)
+		return err;
 
-	return 0;
+	return ret;
 }
 
 static int rv3029_eeprom_update_bits(struct device *dev,
Signature · 1 line
2.20.1
author maintainer
#241 lore raw

PATCH 503/671 mac80211: minstrel_ht: fix per-group max throughput rate initialization

In reply to Sasha Levin

From: Felix Fietkau <[email protected]>

[ Upstream commit 56dd918ff06e3ee24d8067e93ed12b2a39e71394 ]

The group number needs to be multiplied by the number of rates per group to get the full rate index

Fixes 5935839ad735 ("mac80211: improve minstrel_ht rate sorting by throughput & probability")
---
 net/mac80211/rc80211_minstrel_ht.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/mac80211/rc80211_minstrel_ht.c +1 -1
index 3d5520776655..0b60e330c115 100644
@@ -529,7 +529,7 @@ minstrel_ht_update_stats(struct minstrel_priv *mp, struct minstrel_ht_sta *mi)
 
 		/* (re)Initialize group rate indexes */
 		for(j = 0; j < MAX_THR_RATES; j++)
-			tmp_group_tp_rate[j] = group;
+			tmp_group_tp_rate[j] = MCS_GROUP_RATES * group;
 
 		for (i = 0; i < MCS_GROUP_RATES; i++) {
 			if (!(mi->supported[group] & BIT(i)))
Signature · 1 line
2.20.1
author maintainer
#242 lore raw

PATCH 504/671 i40e: reduce stack usage in i40e_set_fc

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 33b165684ab70867d4545643f550a5d48d3ddc57 ]

The functions i40e_aq_get_phy_abilities_resp() and i40e_set_fc() both have giant structure on the stack, which makes each one use stack frames larger than 500 bytes.

As clang decides one function into the other, we get a warning for exceeding the frame size limit on 32-bit architectures:

drivers/net/ethernet/intel/i40e/i40e_common.c:1654:23: error: stack frame size of 1116 bytes in function 'i40e_set_fc' [-Werror,-Wframe-larger-than=]

When building with gcc, the inlining does not happen, but i40e_set_fc() calls i40e_aq_get_phy_abilities_resp() anyway, so they add up on the kernel stack just as much.

The parts that actually use large stacks don't overlap, so make sure each one is a separate function, and mark them as noinline_for_stack to prevent the compilers from combining them again.

Fixes 0a862b43acc6 ("i40e/i40evf: Add module_types and update_link_info")
---
 drivers/net/ethernet/intel/i40e/i40e_common.c | 91 +++++++++++--------
 1 file changed, 51 insertions(+), 40 deletions(-)
drivers/net/ethernet/intel/i40e/i40e_common.c +51 -40
index 85f75b5978fc..eb0ae6ab01e2 100644
@@ -1668,25 +1668,15 @@ enum i40e_status_code i40e_aq_set_phy_config(struct i40e_hw *hw,
 	return status;
 }
 
-/**
- * i40e_set_fc
- * @hw: pointer to the hw struct
- * @aq_failures: buffer to return AdminQ failure information
- * @atomic_restart: whether to enable atomic link restart
- *
- * Set the requested flow control mode using set_phy_config.
- **/
-enum i40e_status_code i40e_set_fc(struct i40e_hw *hw, u8 *aq_failures,
-				  bool atomic_restart)
+static noinline_for_stack enum i40e_status_code
+i40e_set_fc_status(struct i40e_hw *hw,
+		   struct i40e_aq_get_phy_abilities_resp *abilities,
+		   bool atomic_restart)
 {
-	enum i40e_fc_mode fc_mode = hw->fc.requested_mode;
-	struct i40e_aq_get_phy_abilities_resp abilities;
 	struct i40e_aq_set_phy_config config;
-	enum i40e_status_code status;
+	enum i40e_fc_mode fc_mode = hw->fc.requested_mode;
 	u8 pause_mask = 0x0;
 
-	*aq_failures = 0x0;
-
 	switch (fc_mode) {
 	case I40E_FC_FULL:
 		pause_mask |= I40E_AQ_PHY_FLAG_PAUSE_TX;
@@ -1702,6 +1692,48 @@ enum i40e_status_code i40e_set_fc(struct i40e_hw *hw, u8 *aq_failures,
 		break;
 	}
 
+	memset(&config, 0, sizeof(struct i40e_aq_set_phy_config));
+	/* clear the old pause settings */
+	config.abilities = abilities->abilities & ~(I40E_AQ_PHY_FLAG_PAUSE_TX) &
+			   ~(I40E_AQ_PHY_FLAG_PAUSE_RX);
+	/* set the new abilities */
+	config.abilities |= pause_mask;
+	/* If the abilities have changed, then set the new config */
+	if (config.abilities == abilities->abilities)
+		return 0;
+
+	/* Auto restart link so settings take effect */
+	if (atomic_restart)
+		config.abilities |= I40E_AQ_PHY_ENABLE_ATOMIC_LINK;
+	/* Copy over all the old settings */
+	config.phy_type = abilities->phy_type;
+	config.phy_type_ext = abilities->phy_type_ext;
+	config.link_speed = abilities->link_speed;
+	config.eee_capability = abilities->eee_capability;
+	config.eeer = abilities->eeer_val;
+	config.low_power_ctrl = abilities->d3_lpan;
+	config.fec_config = abilities->fec_cfg_curr_mod_ext_info &
+			    I40E_AQ_PHY_FEC_CONFIG_MASK;
+
+	return i40e_aq_set_phy_config(hw, &config, NULL);
+}
+
+/**
+ * i40e_set_fc
+ * @hw: pointer to the hw struct
+ * @aq_failures: buffer to return AdminQ failure information
+ * @atomic_restart: whether to enable atomic link restart
+ *
+ * Set the requested flow control mode using set_phy_config.
+ **/
+enum i40e_status_code i40e_set_fc(struct i40e_hw *hw, u8 *aq_failures,
+				  bool atomic_restart)
+{
+	struct i40e_aq_get_phy_abilities_resp abilities;
+	enum i40e_status_code status;
+
+	*aq_failures = 0x0;
+
 	/* Get the current phy config */
 	status = i40e_aq_get_phy_capabilities(hw, false, false, &abilities,
 					      NULL);
@@ -1710,31 +1742,10 @@ enum i40e_status_code i40e_set_fc(struct i40e_hw *hw, u8 *aq_failures,
 		return status;
 	}
 
-	memset(&config, 0, sizeof(struct i40e_aq_set_phy_config));
-	/* clear the old pause settings */
-	config.abilities = abilities.abilities & ~(I40E_AQ_PHY_FLAG_PAUSE_TX) &
-			   ~(I40E_AQ_PHY_FLAG_PAUSE_RX);
-	/* set the new abilities */
-	config.abilities |= pause_mask;
-	/* If the abilities have changed, then set the new config */
-	if (config.abilities != abilities.abilities) {
-		/* Auto restart link so settings take effect */
-		if (atomic_restart)
-			config.abilities |= I40E_AQ_PHY_ENABLE_ATOMIC_LINK;
-		/* Copy over all the old settings */
-		config.phy_type = abilities.phy_type;
-		config.phy_type_ext = abilities.phy_type_ext;
-		config.link_speed = abilities.link_speed;
-		config.eee_capability = abilities.eee_capability;
-		config.eeer = abilities.eeer_val;
-		config.low_power_ctrl = abilities.d3_lpan;
-		config.fec_config = abilities.fec_cfg_curr_mod_ext_info &
-				    I40E_AQ_PHY_FEC_CONFIG_MASK;
-		status = i40e_aq_set_phy_config(hw, &config, NULL);
+	status = i40e_set_fc_status(hw, &abilities, atomic_restart);
+	if (status)
+		*aq_failures |= I40E_SET_FC_AQ_FAIL_SET;
 
-		if (status)
-			*aq_failures |= I40E_SET_FC_AQ_FAIL_SET;
-	}
 	/* Update the link info */
 	status = i40e_update_link_info(hw);
 	if (status) {
@@ -2563,7 +2574,7 @@ i40e_status i40e_get_link_status(struct i40e_hw *hw, bool *link_up)
  * i40e_updatelink_status - update status of the HW network link
  * @hw: pointer to the hw struct
  **/
-i40e_status i40e_update_link_info(struct i40e_hw *hw)
+noinline_for_stack i40e_status i40e_update_link_info(struct i40e_hw *hw)
 {
 	struct i40e_aq_get_phy_abilities_resp abilities;
 	i40e_status status = 0;
Signature · 1 line
2.20.1
author maintainer
#243 lore raw

PATCH 505/671 media: atmel: atmel-isi: fix timeout value for stop streaming

In reply to Sasha Levin

From: Alexandre Kroupski <[email protected]>

[ Upstream commit 623fd246bb40234fe68dd4e7c1f1f081f9c45a3d ]

In case of sensor malfunction, stop streaming timeout takes much longer than expected. This is due to conversion of time to jiffies: milliseconds multiplied with HZ (ticks/second) gives out a value of jiffies with 10^3 greater. We need to also divide by 10^3 to obtain the right jiffies value. In other words FRAME_INTERVAL_MILLI_SEC must be in seconds in order to multiply by HZ and get the right jiffies value to add to the current jiffies for the timeout expire time.

Fixes 195ebc43bf76 ("[media] V4L: at91: add Atmel Image Sensor Interface (ISI) support")
---
 drivers/media/platform/atmel/atmel-isi.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/atmel/atmel-isi.c +1 -1
index e8db4df1e7c4..1a0e5233ae28 100644
@@ -496,7 +496,7 @@ static void stop_streaming(struct vb2_queue *vq)
 	spin_unlock_irq(&isi->irqlock);
 
 	if (!isi->enable_preview_path) {
-		timeout = jiffies + FRAME_INTERVAL_MILLI_SEC * HZ;
+		timeout = jiffies + (FRAME_INTERVAL_MILLI_SEC * HZ) / 1000;
 		/* Wait until the end of the current frame. */
 		while ((isi_readl(isi, ISI_STATUS) & ISI_CTRL_CDC) &&
 				time_before(jiffies, timeout))
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#244 lore raw

PATCH 506/671 ARM: 8896/1: VDSO: Don't leak kernel addresses

In reply to Sasha Levin

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit 3e07590e7248db951fed6a2039403b5a39010be7 ]

Since commit ad67b74d2469d9b8 ("printk: hash addresses printed with %p"), an obfuscated kernel pointer is printed at every boot if debugging is enabled:

    vdso: 1 text pages at base (____ptrval____)

Remove the print completely, as it's useless without the address.

Based on commit 0f1bf7e39822476b ("arm64/vdso: don't leak kernel addresses").

Fixes ad67b74d2469d9b8 ("printk: hash addresses printed with %p")
---
 arch/arm/kernel/vdso.c | 1 -
 1 file changed, 1 deletion(-)
arch/arm/kernel/vdso.c +0 -1
index f4dd7f9663c1..e8cda5e02b4e 100644
@@ -205,7 +205,6 @@ static int __init vdso_init(void)
 	}
 
 	text_pages = (vdso_end - vdso_start) >> PAGE_SHIFT;
-	pr_debug("vdso: %i text pages at base %p\n", text_pages, vdso_start);
 
 	/* Allocate the VDSO text pagelist */
 	vdso_text_pagelist = kcalloc(text_pages, sizeof(struct page *),
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#245 lore raw

PATCH 507/671 rtc: pcf2127: bugfix: read rtc disables watchdog

In reply to Sasha Levin

From: Bruno Thomsen <[email protected]>

[ Upstream commit 7f43020e3bdb63d65661ed377682702f8b34d3ea ]

The previous fix listed bulk read of registers as root cause of accendential disabling of watchdog, since the watchdog counter register (WD_VAL) was zeroed.

Fixes 3769a375ab83 rtc: pcf2127: bulk read only date and time registers.

Tested with the same PCF2127 chip as Sean reveled root cause of WD_VAL register value zeroing was caused by reading CTRL2 register which is one of the watchdog feature control registers.

So the solution is to not read the first two control registers (CTRL1 and CTRL2) in pcf2127_rtc_read_time as they are not needed anyway. Size of local buf variable is kept to allow easy usage of register defines to improve readability of code.

Debug trace line was updated after CTRL1 and CTRL2 are no longer read from the chip. Also replaced magic numbers in buf access with register defines.

---
 drivers/rtc/rtc-pcf2127.c | 32 ++++++++++++--------------------
 1 file changed, 12 insertions(+), 20 deletions(-)
drivers/rtc/rtc-pcf2127.c +12 -20
index 7cb786d76e3c..8c62406f92dd 100644
@@ -57,20 +57,14 @@ static int pcf2127_rtc_read_time(struct device *dev, struct rtc_time *tm)
 	struct pcf2127 *pcf2127 = dev_get_drvdata(dev);
 	unsigned char buf[10];
 	int ret;
-	int i;
 
-	for (i = 0; i <= PCF2127_REG_CTRL3; i++) {
-		ret = regmap_read(pcf2127->regmap, PCF2127_REG_CTRL1 + i,
-				  (unsigned int *)(buf + i));
-		if (ret) {
-			dev_err(dev, "%s: read error\n", __func__);
-			return ret;
-		}
-	}
-
-	ret = regmap_bulk_read(pcf2127->regmap, PCF2127_REG_SC,
-			       (buf + PCF2127_REG_SC),
-			       ARRAY_SIZE(buf) - PCF2127_REG_SC);
+	/*
+	 * Avoid reading CTRL2 register as it causes WD_VAL register
+	 * value to reset to 0 which means watchdog is stopped.
+	 */
+	ret = regmap_bulk_read(pcf2127->regmap, PCF2127_REG_CTRL3,
+			       (buf + PCF2127_REG_CTRL3),
+			       ARRAY_SIZE(buf) - PCF2127_REG_CTRL3);
 	if (ret) {
 		dev_err(dev, "%s: read error\n", __func__);
 		return ret;
@@ -91,14 +85,12 @@ static int pcf2127_rtc_read_time(struct device *dev, struct rtc_time *tm)
 	}
 
 	dev_dbg(dev,
-		"%s: raw data is cr1=%02x, cr2=%02x, cr3=%02x, "
-		"sec=%02x, min=%02x, hr=%02x, "
+		"%s: raw data is cr3=%02x, sec=%02x, min=%02x, hr=%02x, "
 		"mday=%02x, wday=%02x, mon=%02x, year=%02x\n",
-		__func__,
-		buf[0], buf[1], buf[2],
-		buf[3], buf[4], buf[5],
-		buf[6], buf[7], buf[8], buf[9]);
-
+		__func__, buf[PCF2127_REG_CTRL3], buf[PCF2127_REG_SC],
+		buf[PCF2127_REG_MN], buf[PCF2127_REG_HR],
+		buf[PCF2127_REG_DM], buf[PCF2127_REG_DW],
+		buf[PCF2127_REG_MO], buf[PCF2127_REG_YR]);
 
 	tm->tm_sec = bcd2bin(buf[PCF2127_REG_SC] & 0x7F);
 	tm->tm_min = bcd2bin(buf[PCF2127_REG_MN] & 0x7F);
Signature · 1 line
2.20.1
author maintainer
#246 lore raw

PATCH 508/671 mips: avoid explicit UB in assignment of mips_io_port_base

In reply to Sasha Levin

From: Nick Desaulniers <[email protected]>

[ Upstream commit 12051b318bc3ce5b42d6d786191008284b067d83 ]

The code in question is modifying a variable declared const through pointer manipulation. Such code is explicitly undefined behavior, and is the lone issue preventing malta_defconfig from booting when built with Clang:

If an attempt is made to modify an object defined with a const-qualified type through use of an lvalue with non-const-qualified type, the behavior is undefined.

LLVM is removing such assignments. A simple fix is to not declare variables const that you plan on modifying. Limiting the scope would be a better method of preventing unwanted writes to such a variable.

Further, the code in question mentions "compiler bugs" without any links to bug reports, so it is difficult to know if the issue is resolved in GCC. The patch was authored in 2006, which would have been GCC 4.0.3 or 4.1.1. The minimal supported version of GCC in the Linux kernel is currently 4.6.

For what its worth, there was UB before the commit in question, it just added a barrier and got lucky IRT codegen. I don't think there's any actual compiler bugs related, just runtime bugs due to UB.

---
 arch/mips/include/asm/io.h | 14 ++------------
 arch/mips/kernel/setup.c   |  2 +-
 2 files changed, 3 insertions(+), 13 deletions(-)
arch/mips/include/asm/io.h +2 -12
index 54c730aed327..df1eaa365279 100644
@@ -62,21 +62,11 @@
  * instruction, so the lower 16 bits must be zero.  Should be true on
  * on any sane architecture; generic code does not use this assumption.
  */
-extern const unsigned long mips_io_port_base;
+extern unsigned long mips_io_port_base;
 
-/*
- * Gcc will generate code to load the value of mips_io_port_base after each
- * function call which may be fairly wasteful in some cases.  So we don't
- * play quite by the book.  We tell gcc mips_io_port_base is a long variable
- * which solves the code generation issue.  Now we need to violate the
- * aliasing rules a little to make initialization possible and finally we
- * will need the barrier() to fight side effects of the aliasing chat.
- * This trickery will eventually collapse under gcc's optimizer.  Oh well.
- */
 static inline void set_io_port_base(unsigned long base)
 {
-	* (unsigned long *) &mips_io_port_base = base;
-	barrier();
+	mips_io_port_base = base;
 }
 
 /*
arch/mips/kernel/setup.c +1 -1
index 8aaaa42f91ed..e87c98b8a72c 100644
@@ -76,7 +76,7 @@ static char __initdata builtin_cmdline[COMMAND_LINE_SIZE] = CONFIG_CMDLINE;
  * mips_io_port_base is the begin of the address space to which x86 style
  * I/O ports are mapped.
  */
-const unsigned long mips_io_port_base = -1;
+unsigned long mips_io_port_base = -1;
 EXPORT_SYMBOL(mips_io_port_base);
 
 static struct resource code_resource = { .name = "Kernel code", };
Signature · 1 line
2.20.1
author maintainer
#247 lore raw

PATCH 509/671 media: em28xx: Fix exception handling in em28xx_alloc_urbs()

In reply to Sasha Levin

From: Markus Elfring <[email protected]>

[ Upstream commit ecbce48f1ff2442371ebcd12ec0ecddb431fbd72 ]

A null pointer would be passed to a call of the function "kfree" directly after a call of the function "kcalloc" failed at one place. Pass the data structure member "urb" instead for which memory was allocated before (so that this resource will be properly cleaned up).

This issue was detected by using the Coccinelle software.

Fixes d571b592c6206d33731f41aa710fa0f69ac8611b ("media: em28xx: don't use coherent buffer for DMA transfers")
---
 drivers/media/usb/em28xx/em28xx-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/usb/em28xx/em28xx-core.c +1 -1
index 5657f8710ca6..69445c8e38e2 100644
@@ -930,7 +930,7 @@ int em28xx_alloc_urbs(struct em28xx *dev, enum em28xx_mode mode, int xfer_bulk,
 
 	usb_bufs->buf = kcalloc(num_bufs, sizeof(void *), GFP_KERNEL);
 	if (!usb_bufs->buf) {
-		kfree(usb_bufs->buf);
+		kfree(usb_bufs->urb);
 		return -ENOMEM;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#248 lore raw

PATCH 510/671 iommu/mediatek: Fix iova_to_phys PA start for 4GB mode

In reply to Sasha Levin

From: Yong Wu <[email protected]>

[ Upstream commit 76ce65464fcd2c21db84391572b7938b716aceb0 ]

In M4U 4GB mode, the physical address is remapped as below:

CPU Physical address:

====================

0      1G       2G     3G       4G     5G
|---A---|---B---|---C---|---D---|---E---|

+--I/O--+------------Memory-------------+

IOMMU output physical address:
=============================

                                4G      5G     6G      7G      8G
                                |---E---|---B---|---C---|---D---|
                                +------------Memory-------------+

The Region 'A'(I/O) can not be mapped by M4U; For Region 'B'/'C'/'D', the bit32 of the CPU physical address always is needed to set, and for Region 'E', the CPU physical address keep as is. something looks like this:

CPU PA         ->    M4U OUTPUT PA
0x4000_0000          0x1_4000_0000 (Add bit32)
0x8000_0000          0x1_8000_0000 ...
0xc000_0000          0x1_c000_0000 ...
0x1_0000_0000        0x1_0000_0000 (No change)

Additionally, the iommu consumers always use the CPU phyiscal address.

The PA in the iova_to_phys that is got from v7s always is u32, But from the CPU point of view, PA only need add BIT(32) when PA < 0x4000_0000.

Fixes 30e2fccf9512 ("iommu/mediatek: Enlarge the validate PA range

for 4GB mode")

---
 drivers/iommu/mtk_iommu.c | 26 +++++++++++++++++++++++++-
 1 file changed, 25 insertions(+), 1 deletion(-)
drivers/iommu/mtk_iommu.c +25 -1
index f9f69f7111a9..c88cc299e3c9 100644
@@ -115,6 +115,30 @@ struct mtk_iommu_domain {
 
 static struct iommu_ops mtk_iommu_ops;
 
+/*
+ * In M4U 4GB mode, the physical address is remapped as below:
+ *
+ * CPU Physical address:
+ * ====================
+ *
+ * 0      1G       2G     3G       4G     5G
+ * |---A---|---B---|---C---|---D---|---E---|
+ * +--I/O--+------------Memory-------------+
+ *
+ * IOMMU output physical address:
+ *  =============================
+ *
+ *                                 4G      5G     6G      7G      8G
+ *                                 |---E---|---B---|---C---|---D---|
+ *                                 +------------Memory-------------+
+ *
+ * The Region 'A'(I/O) can NOT be mapped by M4U; For Region 'B'/'C'/'D', the
+ * bit32 of the CPU physical address always is needed to set, and for Region
+ * 'E', the CPU physical address keep as is.
+ * Additionally, The iommu consumers always use the CPU phyiscal address.
+ */
+#define MTK_IOMMU_4GB_MODE_REMAP_BASE	 0x40000000
+
 static LIST_HEAD(m4ulist);	/* List all the M4U HWs */
 
 #define for_each_m4u(data)	list_for_each_entry(data, &m4ulist, list)
@@ -409,7 +433,7 @@ static phys_addr_t mtk_iommu_iova_to_phys(struct iommu_domain *domain,
 	pa = dom->iop->iova_to_phys(dom->iop, iova);
 	spin_unlock_irqrestore(&dom->pgtlock, flags);
 
-	if (data->enable_4GB)
+	if (data->enable_4GB && pa < MTK_IOMMU_4GB_MODE_REMAP_BASE)
 		pa |= BIT_ULL(32);
 
 	return pa;
Signature · 6 lines
2.20.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author maintainer
#249 lore raw

PATCH 511/671 ahci: Do not export local variable ahci_em_messages

In reply to Sasha Levin

From: Andy Shevchenko <[email protected]>

[ Upstream commit 60fc35f327e0a9e60b955c0f3c3ed623608d1baa ]

The commit ed08d40cdec4
("ahci: Changing two module params with static and __read_mostly")
moved ahci_em_messages to be static while missing the fact of exporting it.

WARNING: "ahci_em_messages" [vmlinux] is a static EXPORT_SYMBOL_GPL

Drop export for the local variable ahci_em_messages.

Fixes ed08d40cdec4 ("ahci: Changing two module params with static and __read_mostly")
Cc Chuansheng Liu <[email protected]>
---
 drivers/ata/libahci.c | 1 -
 1 file changed, 1 deletion(-)
drivers/ata/libahci.c +0 -1
index b5f57c69c487..2bdb250a2142 100644
@@ -191,7 +191,6 @@ struct ata_port_operations ahci_pmp_retry_srst_ops = {
 EXPORT_SYMBOL_GPL(ahci_pmp_retry_srst_ops);
 
 static bool ahci_em_messages __read_mostly = true;
-EXPORT_SYMBOL_GPL(ahci_em_messages);
 module_param(ahci_em_messages, bool, 0444);
 /* add other LED protocol types when they become supported */
 MODULE_PARM_DESC(ahci_em_messages,
Signature · 1 line
2.20.1
author maintainer
#250 lore raw

PATCH 512/671 rxrpc: Fix lack of conn cleanup when local endpoint is cleaned up [ver #2]

In reply to Sasha Levin

From: David Howells <[email protected]>

[ Upstream commit d12040b6933f684a26773afad46dbba9778608d7 ]

When a local endpoint is ceases to be in use, such as when the kafs module is unloaded, the kernel will emit an assertion failure if there are any outstanding client connections:

rxrpc: Assertion failed ------------[ cut here ]------------ kernel BUG at net/rxrpc/local_object.c:433!

and even beyond that, will evince other oopses if there are service connections still present.

Fix this by:

(1) Removing the triggering of connection reaping when an rxrpc socket is released. These don't actually clean up the connections anyway - and further, the local endpoint may still be in use through another socket.

(2) Mark the local endpoint as dead when we start the process of tearing it down.

(3) When destroying a local endpoint, strip all of its client connections from the idle list and discard the ref on each that the list was holding.

(4) When destroying a local endpoint, call the service connection reaper directly (rather than through a workqueue) to immediately kill off all outstanding service connections.

(5) Make the service connection reaper reap connections for which the local endpoint is marked dead.

Only after destroying the connections can we close the socket lest we get an oops in a workqueue that's looking at a connection or a peer.

Fixes 3d18cbb7fd0c ("rxrpc: Fix conn expiry timers")
---
 net/rxrpc/af_rxrpc.c     |  3 ---
 net/rxrpc/ar-internal.h  |  1 +
 net/rxrpc/conn_client.c  | 44 ++++++++++++++++++++++++++++++++++++++++
 net/rxrpc/conn_object.c  |  2 +-
 net/rxrpc/local_object.c |  5 ++++-
 5 files changed, 50 insertions(+), 5 deletions(-)
net/rxrpc/af_rxrpc.c +0 -3
index 7319d3ca30e9..a74edb10cbfc 100644
@@ -869,7 +869,6 @@ static void rxrpc_sock_destructor(struct sock *sk)
 static int rxrpc_release_sock(struct sock *sk)
 {
 	struct rxrpc_sock *rx = rxrpc_sk(sk);
-	struct rxrpc_net *rxnet = rxrpc_net(sock_net(&rx->sk));
 
 	_enter("%p{%d,%d}", sk, sk->sk_state, refcount_read(&sk->sk_refcnt));
 
@@ -905,8 +904,6 @@ static int rxrpc_release_sock(struct sock *sk)
 	rxrpc_release_calls_on_socket(rx);
 	flush_workqueue(rxrpc_workqueue);
 	rxrpc_purge_queue(&sk->sk_receive_queue);
-	rxrpc_queue_work(&rxnet->service_conn_reaper);
-	rxrpc_queue_work(&rxnet->client_conn_reaper);
 
 	rxrpc_unuse_local(rx->local);
 	rx->local = NULL;
net/rxrpc/ar-internal.h +1 -0
index 37e4c1801a41..ccef6e40e002 100644
@@ -903,6 +903,7 @@ void rxrpc_disconnect_client_call(struct rxrpc_call *);
 void rxrpc_put_client_conn(struct rxrpc_connection *);
 void rxrpc_discard_expired_client_conns(struct work_struct *);
 void rxrpc_destroy_all_client_connections(struct rxrpc_net *);
+void rxrpc_clean_up_local_conns(struct rxrpc_local *);
 
 /*
  * conn_event.c
net/rxrpc/conn_client.c +44 -0

47 lines not on this page · raw message

net/rxrpc/conn_object.c +1 -1
index 885dae829f4a..004a6eb529bc 100644
@@ -401,7 +401,7 @@ void rxrpc_service_connection_reaper(struct work_struct *work)
 		if (conn->state == RXRPC_CONN_SERVICE_PREALLOC)
 			continue;
 
-		if (rxnet->live) {
+		if (rxnet->live && !conn->params.local->dead) {
 			idle_timestamp = READ_ONCE(conn->idle_timestamp);
 			expire_at = idle_timestamp + rxrpc_connection_expiry * HZ;
 			if (conn->params.local->service_closed)
net/rxrpc/local_object.c +4 -1
index c752ad487067..04f0976841a4 100644
@@ -430,11 +430,14 @@ static void rxrpc_local_destroyer(struct rxrpc_local *local)
 
 	_enter("%d", local->debug_id);
 
+	local->dead = true;
+
 	mutex_lock(&rxnet->local_mutex);
 	list_del_init(&local->link);
 	mutex_unlock(&rxnet->local_mutex);
 
-	ASSERT(RB_EMPTY_ROOT(&local->client_conns));
+	rxrpc_clean_up_local_conns(local);
+	rxrpc_service_connection_reaper(&rxnet->service_conn_reaper);
 	ASSERT(!local->service);
 
 	if (socket) {
Signature · 1 line
2.20.1
author maintainer
#251 lore raw

PATCH 513/671 Partially revert "kfifo: fix kfifo_alloc() and kfifo_init()"

In reply to Sasha Levin

From: Linus Torvalds <[email protected]>

[ Upstream commit ab9bb6318b0967671e0c9b6537c1537d51ca4f45 ]

Commit dfe2a77fd243 ("kfifo: fix kfifo_alloc() and kfifo_init()") made the kfifo code round the number of elements up. That was good for __kfifo_alloc(), but it's actually wrong for __kfifo_init().

The difference? __kfifo_alloc() will allocate the rounded-up number of elements, but __kfifo_init() uses an allocation done by the caller. We can't just say "use more elements than the caller allocated", and have to round down.

The good news? All the normal cases will be using power-of-two arrays anyway, and most users of kfifo's don't use kfifo_init() at all, but one of the helper macros to declare a KFIFO that enforce the proper power-of-two behavior. But it looks like at least ibmvscsis might be affected.

The bad news? Will Deacon refers to an old thread and points points out that the memory ordering in kfifo's is questionable. See

https://lore.kernel.org/lkml/[email protected]/

for more.

Fixes dfe2a77fd243 ("kfifo: fix kfifo_alloc() and kfifo_init()")
Cc Stefani Seibold <[email protected]>
Cc Andrew Morton <[email protected]>
Cc Dan Carpenter <[email protected]>
Cc Greg KH <[email protected]>
Cc Kees Cook <[email protected]>
Cc Will Deacon <[email protected]>
---
 lib/kfifo.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
lib/kfifo.c +2 -1
index 015656aa8182..6320ab91e343 100644
@@ -82,7 +82,8 @@ int __kfifo_init(struct __kfifo *fifo, void *buffer,
 {
 	size /= esize;
 
-	size = roundup_pow_of_two(size);
+	if (!is_power_of_2(size))
+		size = rounddown_pow_of_two(size);
 
 	fifo->in = 0;
 	fifo->out = 0;
Signature · 1 line
2.20.1
author maintainer
#252 lore raw

PATCH 514/671 hwmon: (lm75) Fix write operations for negative temperatures

In reply to Sasha Levin

From: Guenter Roeck <[email protected]>

[ Upstream commit 7d82fcc9d9e81241778aaa22fda7be753e237d86 ]

Writes into limit registers fail if the temperature written is negative. The regmap write operation checks the value range, regmap_write accepts an unsigned int as parameter, and the temperature value passed to regmap_write is kept in a variable declared as long. Negative values are converted large unsigned integers, which fails the range check. Fix by type casting the temperature to u16 when calling regmap_write().

Cc Iker Perez del Palomar Sustatxa <[email protected]>
Fixes e65365fed87f ("hwmon: (lm75) Convert to use regmap")
---
 drivers/hwmon/lm75.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hwmon/lm75.c +1 -1
index 49f4b33a5685..7f28912c9abc 100644
@@ -165,7 +165,7 @@ static int lm75_write(struct device *dev, enum hwmon_sensor_types type,
 	temp = DIV_ROUND_CLOSEST(temp  << (resolution - 8),
 				 1000) << (16 - resolution);
 
-	return regmap_write(data->regmap, reg, temp);
+	return regmap_write(data->regmap, reg, (u16)temp);
 }
 
 static umode_t lm75_is_visible(const void *data, enum hwmon_sensor_types type,
Signature · 1 line
2.20.1
author maintainer
#253 lore raw

PATCH 515/671 net/sched: cbs: Set default link speed to 10 Mbps in cbs_set_port_rate

In reply to Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 1c6c09a0ae62fa3ea8f8ead2ac3920e6fff2de64 ]

The discussion to be made is absolutely the same as in the case of previous patch ("taprio: Set default link speed to 10 Mbps in taprio_set_picos_per_byte"). Nothing is lost when setting a default.

Cc Leandro Dorileo <[email protected]>
Fixes e0a7683d30e9 ("net/sched: cbs: fix port_rate miscalculation")
---
 net/sched/sch_cbs.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)
net/sched/sch_cbs.c +11 -8
index b3c8d04929df..289f66b9238d 100644
@@ -185,11 +185,6 @@ static struct sk_buff *cbs_dequeue_soft(struct Qdisc *sch)
 	s64 credits;
 	int len;
 
-	if (atomic64_read(&q->port_rate) == -1) {
-		WARN_ONCE(1, "cbs: dequeue() called with unknown port rate.");
-		return NULL;
-	}
-
 	if (q->credits < 0) {
 		credits = timediff_to_credits(now - q->last, q->idleslope);
 
@@ -307,11 +302,19 @@ static int cbs_enable_offload(struct net_device *dev, struct cbs_sched_data *q,
 static void cbs_set_port_rate(struct net_device *dev, struct cbs_sched_data *q)
 {
 	struct ethtool_link_ksettings ecmd;
+	int speed = SPEED_10;
 	int port_rate = -1;
+	int err;
+
+	err = __ethtool_get_link_ksettings(dev, &ecmd);
+	if (err < 0)
+		goto skip;
+
+	if (ecmd.base.speed != SPEED_UNKNOWN)
+		speed = ecmd.base.speed;
 
-	if (!__ethtool_get_link_ksettings(dev, &ecmd) &&
-	    ecmd.base.speed != SPEED_UNKNOWN)
-		port_rate = ecmd.base.speed * 1000 * BYTES_PER_KBIT;
+skip:
+	port_rate = speed * 1000 * BYTES_PER_KBIT;
 
 	atomic64_set(&q->port_rate, port_rate);
 	netdev_dbg(dev, "cbs: set %s's port_rate to: %lld, linkspeed: %d\n",
Signature · 1 line
2.20.1
author maintainer
#254 lore raw

PATCH 516/671 power: supply: Init device wakeup after device_add()

In reply to Sasha Levin

From: Stephen Boyd <[email protected]>

[ Upstream commit 8288022284859acbcc3cf1a073a1e2692d6c2543 ]

We may want to use the device pointer in device_init_wakeup() with functions that expect the device to already be added with device_add(). For example, if we were to link the device initializing wakeup to something in sysfs such as a class for wakeups we'll run into an error. It looks like this code was written with the assumption that the device would be added before initializing wakeup due to the order of operations in power_supply_unregister().

Let's change the order of operations so we don't run into problems here.

Fixes 948dcf966228 ("power_supply: Prevent suspend until power supply events are processed")
Cc Greg Kroah-Hartman <[email protected]>
Cc Tri Vo <[email protected]>
Cc Kalesh Singh <[email protected]>
Cc Ravi Chandra Sadineni <[email protected]>
Cc Viresh Kumar <[email protected]>
---
 drivers/power/supply/power_supply_core.c | 10 +++++-----
 1 file changed, 5 insertions(+), 5 deletions(-)
drivers/power/supply/power_supply_core.c +5 -5
index e85361878450..e43a7b3b570c 100644
@@ -902,14 +902,14 @@ __power_supply_register(struct device *parent,
 	}
 
 	spin_lock_init(&psy->changed_lock);
-	rc = device_init_wakeup(dev, ws);
-	if (rc)
-		goto wakeup_init_failed;
-
 	rc = device_add(dev);
 	if (rc)
 		goto device_add_failed;
 
+	rc = device_init_wakeup(dev, ws);
+	if (rc)
+		goto wakeup_init_failed;
+
 	rc = psy_register_thermal(psy);
 	if (rc)
 		goto register_thermal_failed;
@@ -946,8 +946,8 @@ __power_supply_register(struct device *parent,
 	psy_unregister_thermal(psy);
 register_thermal_failed:
 	device_del(dev);
-device_add_failed:
 wakeup_init_failed:
+device_add_failed:
 check_supplies_failed:
 dev_set_name_failed:
 	put_device(dev);
Signature · 1 line
2.20.1
author maintainer
#255 lore raw

PATCH 517/671 x86, perf: Fix the dependency of the x86 insn decoder selftest

In reply to Sasha Levin

From: Masami Hiramatsu <[email protected]>

[ Upstream commit 7720804a2ae46c90265a32c81c45fb6f8d2f4e8b ]

Since x86 instruction decoder is not only for kprobes, it should be tested when the insn.c is compiled. (e.g. perf is enabled but kprobes is disabled)

Cc Linus Torvalds <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Thomas Gleixner <[email protected]>
Fixes cbe5c34c8c1f ("x86: Compile insn.c and inat.c only for KPROBES")
---
 arch/x86/Kconfig.debug | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/Kconfig.debug +1 -1
index 7d68f0c7cfb1..687cd1a213d5 100644
@@ -181,7 +181,7 @@ config HAVE_MMIOTRACE_SUPPORT
 
 config X86_DECODER_SELFTEST
 	bool "x86 instruction decoder selftest"
-	depends on DEBUG_KERNEL && KPROBES
+	depends on DEBUG_KERNEL && INSTRUCTION_DECODER
 	depends on !COMPILE_TEST
 	---help---
 	 Perform x86 instruction decoder selftests at build time.
Signature · 1 line
2.20.1
author maintainer
#256 lore raw

PATCH 518/671 staging: greybus: light: fix a couple double frees

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 329101244f214952606359d254ae883b7109e1a5 ]

The problem is in gb_lights_request_handler(). If we get a request to change the config then we release the light with gb_lights_light_release() and re-allocated it. However, if the allocation fails part way through then we call gb_lights_light_release() again. This can lead to a couple different double frees where we haven't cleared out the original values:

	gb_lights_light_v4l2_unregister(light);
	...
	kfree(light->channels);
	kfree(light->name);

I also made a small change to how we set "light->channels_count = 0;". The original code handled this part fine and did not cause a use after free but it was sort of complicated to read.

---
 drivers/staging/greybus/light.c | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/staging/greybus/light.c +6 -6
index 010ae1e9c7fb..40680eaf3974 100644
@@ -1098,21 +1098,21 @@ static void gb_lights_channel_release(struct gb_channel *channel)
 static void gb_lights_light_release(struct gb_light *light)
 {
 	int i;
-	int count;
 
 	light->ready = false;
 
-	count = light->channels_count;
-
 	if (light->has_flash)
 		gb_lights_light_v4l2_unregister(light);
+	light->has_flash = false;
 
-	for (i = 0; i < count; i++) {
+	for (i = 0; i < light->channels_count; i++)
 		gb_lights_channel_release(&light->channels[i]);
-		light->channels_count--;
-	}
+	light->channels_count = 0;
+
 	kfree(light->channels);
+	light->channels = NULL;
 	kfree(light->name);
+	light->name = NULL;
 }
 
 static void gb_lights_release(struct gb_lights *glights)
Signature · 1 line
2.20.1
author maintainer
#257 lore raw

PATCH 519/671 irqdomain: Add the missing assignment of domain->fwnode for named fwnode

In reply to Sasha Levin

From: Dexuan Cui <[email protected]>

[ Upstream commit 711419e504ebd68c8f03656616829c8ad7829389 ]

Recently device pass-through stops working for Linux VM running on Hyper-V.

git-bisect shows the regression is caused by the recent commit 467a3bb97432 ("PCI: hv: Allocate a named fwnode ..."), but the root cause is that the commit d59f6617eef0 forgets to set the domain->fwnode for IRQCHIP_FWNODE_NAMED*, and as a result:

1. The domain->fwnode remains to be NULL.

2. irq_find_matching_fwspec() returns NULL since "h->fwnode == fwnode" is false, and pci_set_bus_msi_domain() sets the Hyper-V PCI root bus's msi_domain to NULL.

3. When the device is added onto the root bus, the device's dev->msi_domain is set to NULL in pci_set_msi_domain().

4. When a device driver tries to enable MSI-X, pci_msi_setup_msi_irqs() calls arch_setup_msi_irqs(), which uses the native MSI chip (i.e. arch/x86/kernel/apic/msi.c: pci_msi_controller) to set up the irqs, but actually pci_msi_setup_msi_irqs() is supposed to call msi_domain_alloc_irqs() with the hbus->irq_domain, which is created in hv_pcie_init_irq_domain() and is associated with the Hyper-V chip hv_msi_irq_chip. Consequently, the irq line is not properly set up, and the device driver can not receive any interrupt.

Fixes d59f6617eef0 ("genirq: Allow fwnode to carry name information only")
Fixes 467a3bb97432 ("PCI: hv: Allocate a named fwnode instead of an address-based one")
---
 kernel/irq/irqdomain.c | 1 +
 1 file changed, 1 insertion(+)
kernel/irq/irqdomain.c +1 -0
index 0b90be360724..6e8520a81dd8 100644
@@ -148,6 +148,7 @@ struct irq_domain *__irq_domain_add(struct fwnode_handle *fwnode, int size,
 		switch (fwid->type) {
 		case IRQCHIP_FWNODE_NAMED:
 		case IRQCHIP_FWNODE_NAMED_ID:
+			domain->fwnode = fwnode;
 			domain->name = kstrdup(fwid->name, GFP_KERNEL);
 			if (!domain->name) {
 				kfree(domain);
Signature · 1 line
2.20.1
author maintainer
#258 lore raw

PATCH 520/671 bcma: fix incorrect update of BCMA_CORE_PCI_MDIO_DATA

In reply to Sasha Levin

From: Colin Ian King <[email protected]>

[ Upstream commit 420c20be08a4597404d272ae9793b642401146eb ]

An earlier commit re-worked the setting of the bitmask and is now assigning v with some bit flags rather than bitwise or-ing them into v, consequently the earlier bit-settings of v are being lost. Fix this by replacing an assignment with the bitwise or instead.

Addresses-Coverity: ("Unused value")

Fixes 2be25cac8402 ("bcma: add constants for PCI and use them")
---
 drivers/bcma/driver_pci.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/bcma/driver_pci.c +2 -2
index f499a469e66d..12b2cc9a3fbe 100644
@@ -78,7 +78,7 @@ static u16 bcma_pcie_mdio_read(struct bcma_drv_pci *pc, u16 device, u8 address)
 		v |= (address << BCMA_CORE_PCI_MDIODATA_REGADDR_SHF_OLD);
 	}
 
-	v = BCMA_CORE_PCI_MDIODATA_START;
+	v |= BCMA_CORE_PCI_MDIODATA_START;
 	v |= BCMA_CORE_PCI_MDIODATA_READ;
 	v |= BCMA_CORE_PCI_MDIODATA_TA;
 
@@ -121,7 +121,7 @@ static void bcma_pcie_mdio_write(struct bcma_drv_pci *pc, u16 device,
 		v |= (address << BCMA_CORE_PCI_MDIODATA_REGADDR_SHF_OLD);
 	}
 
-	v = BCMA_CORE_PCI_MDIODATA_START;
+	v |= BCMA_CORE_PCI_MDIODATA_START;
 	v |= BCMA_CORE_PCI_MDIODATA_WRITE;
 	v |= BCMA_CORE_PCI_MDIODATA_TA;
 	v |= data;
Signature · 1 line
2.20.1
author maintainer
#259 lore raw

PATCH 521/671 usb: typec: tps6598x: Fix build error without CONFIG_REGMAP_I2C

In reply to Sasha Levin

From: YueHaibing <[email protected]>

[ Upstream commit 35af2445dc306403254a181507b390ec9eb725d5 ]

If CONFIG_REGMAP_I2C is not set, building fails:

drivers/usb/typec/tps6598x.o: In function `tps6598x_probe':
tps6598x.c:(.text+0x5f0): undefined reference to `__devm_regmap_init_i2c'

Select REGMAP_I2C to fix this.

Fixes 0a4c005bd171 ("usb: typec: driver for TI TPS6598x USB Power Delivery controllers")
---
 drivers/usb/typec/Kconfig | 1 +
 1 file changed, 1 insertion(+)
drivers/usb/typec/Kconfig +1 -0
index 00878c386dd0..8445890accdf 100644
@@ -95,6 +95,7 @@ source "drivers/usb/typec/ucsi/Kconfig"
 config TYPEC_TPS6598X
 	tristate "TI TPS6598x USB Power Delivery controller driver"
 	depends on I2C
+	select REGMAP_I2C
 	help
 	  Say Y or M here if your system has TI TPS65982 or TPS65983 USB Power
 	  Delivery controller.
Signature · 1 line
2.20.1
author maintainer
#260 lore raw

PATCH 522/671 bcache: Fix an error code in bch_dump_read()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit d66c9920c0cf984cf99cab5036fd5f3a1b7fba46 ]

The copy_to_user() function returns the number of bytes remaining to be copied, but the intention here was to return -EFAULT if the copy fails.

Fixes cafe56359144 ("bcache: A block layer cache")
---
 drivers/md/bcache/debug.c | 5 ++---
 1 file changed, 2 insertions(+), 3 deletions(-)
drivers/md/bcache/debug.c +2 -3
index 8c53d874ada4..f6b60d5908f7 100644
@@ -178,10 +178,9 @@ static ssize_t bch_dump_read(struct file *file, char __user *buf,
 	while (size) {
 		struct keybuf_key *w;
 		unsigned int bytes = min(i->bytes, size);
-		int err = copy_to_user(buf, i->buf, bytes);
 
-		if (err)
-			return err;
+		if (copy_to_user(buf, i->buf, bytes))
+			return -EFAULT;
 
 		ret	 += bytes;
 		buf	 += bytes;
Signature · 1 line
2.20.1
author maintainer
#261 lore raw

PATCH 523/671 iio: dac: ad5380: fix incorrect assignment to val

In reply to Sasha Levin

From: Colin Ian King <[email protected]>

[ Upstream commit b1e18768ef1214c0a8048327918a182cabe09f9d ]

Currently the pointer val is being incorrectly incremented instead of the value pointed to by val. Fix this by adding in the missing * indirection operator.

Addresses-Coverity: ("Unused value")

Fixes c03f2c536818 ("staging:iio:dac: Add AD5380 driver")
---
 drivers/iio/dac/ad5380.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/iio/dac/ad5380.c +1 -1
index 873c2bf637c0..617c9f7fe59a 100644
@@ -221,7 +221,7 @@ static int ad5380_read_raw(struct iio_dev *indio_dev,
 		if (ret)
 			return ret;
 		*val >>= chan->scan_type.shift;
-		val -= (1 << chan->scan_type.realbits) / 2;
+		*val -= (1 << chan->scan_type.realbits) / 2;
 		return IIO_VAL_INT;
 	case IIO_CHAN_INFO_SCALE:
 		*val = 2 * st->vref;
Signature · 1 line
2.20.1
author maintainer
#262 lore raw

PATCH 524/671 netfilter: ctnetlink: honor IPS_OFFLOAD flag

In reply to Sasha Levin

From: Pablo Neira Ayuso <[email protected]>

[ Upstream commit b067fa009c884401d23846251031c1f14d8a9c77 ]

If this flag is set, timeout and state are irrelevant to userspace.

Fixes 90964016e5d3 ("netfilter: nf_conntrack: add IPS_OFFLOAD status bit")
---
 net/netfilter/nf_conntrack_netlink.c | 7 +++++--
 1 file changed, 5 insertions(+), 2 deletions(-)
net/netfilter/nf_conntrack_netlink.c +5 -2
index 7ba9ea55816a..31fa94064a62 100644
@@ -555,10 +555,8 @@ ctnetlink_fill_info(struct sk_buff *skb, u32 portid, u32 seq, u32 type,
 		goto nla_put_failure;
 
 	if (ctnetlink_dump_status(skb, ct) < 0 ||
-	    ctnetlink_dump_timeout(skb, ct) < 0 ||
 	    ctnetlink_dump_acct(skb, ct, type) < 0 ||
 	    ctnetlink_dump_timestamp(skb, ct) < 0 ||
-	    ctnetlink_dump_protoinfo(skb, ct) < 0 ||
 	    ctnetlink_dump_helpinfo(skb, ct) < 0 ||
 	    ctnetlink_dump_mark(skb, ct) < 0 ||
 	    ctnetlink_dump_secctx(skb, ct) < 0 ||
@@ -570,6 +568,11 @@ ctnetlink_fill_info(struct sk_buff *skb, u32 portid, u32 seq, u32 type,
 	    ctnetlink_dump_ct_synproxy(skb, ct) < 0)
 		goto nla_put_failure;
 
+	if (!test_bit(IPS_OFFLOAD_BIT, &ct->status) &&
+	    (ctnetlink_dump_timeout(skb, ct) < 0 ||
+	     ctnetlink_dump_protoinfo(skb, ct) < 0))
+		goto nla_put_failure;
+
 	nlmsg_end(skb, nlh);
 	return skb->len;
 
Signature · 1 line
2.20.1
author maintainer
#263 lore raw

PATCH 525/671 ath9k: dynack: fix possible deadlock in ath_dynack_node_{de}init

In reply to Sasha Levin

From: Lorenzo Bianconi <[email protected]>

[ Upstream commit e1aa1a1db3b01c9890e82cf065cee99962ba1ed9 ]

Fix following lockdep warning disabling bh in
ath_dynack_node_init/ath_dynack_node_deinit

[   75.955878] --------------------------------
[   75.955880] inconsistent {SOFTIRQ-ON-W} -> {IN-SOFTIRQ-W} usage.
[   75.955884] swapper/0/0 [HC0[0]:SC1[3]:HE1:SE0] takes:
[   75.955888] 00000000792a7ee0 (&(&da->qlock)->rlock){+.?.}, at: ath_dynack_sample_ack_ts+0x4d/0xa0 [ath9k_hw]
[   75.955905] {SOFTIRQ-ON-W} state was registered at:
[   75.955912]   lock_acquire+0x9a/0x160
[   75.955917]   _raw_spin_lock+0x2c/0x70
[   75.955927]   ath_dynack_node_init+0x2a/0x60 [ath9k_hw]
[   75.955934]   ath9k_sta_state+0xec/0x160 [ath9k]
[   75.955976]   drv_sta_state+0xb2/0x740 [mac80211]
[   75.956008]   sta_info_insert_finish+0x21a/0x420 [mac80211]
[   75.956039]   sta_info_insert_rcu+0x12b/0x2c0 [mac80211]
[   75.956069]   sta_info_insert+0x7/0x70 [mac80211]
[   75.956093]   ieee80211_prep_connection+0x42e/0x730 [mac80211]
[   75.956120]   ieee80211_mgd_auth.cold+0xb9/0x15c [mac80211]
[   75.956152]   cfg80211_mlme_auth+0x143/0x350 [cfg80211]
[   75.956169]   nl80211_authenticate+0x25e/0x2b0 [cfg80211]
[   75.956172]   genl_family_rcv_msg+0x198/0x400
[   75.956174]   genl_rcv_msg+0x42/0x90
[   75.956176]   netlink_rcv_skb+0x35/0xf0
[   75.956178]   genl_rcv+0x1f/0x30
[   75.956180]   netlink_unicast+0x154/0x200
[   75.956182]   netlink_sendmsg+0x1bf/0x3d0
[   75.956186]   ___sys_sendmsg+0x2c2/0x2f0
[   75.956187]   __sys_sendmsg+0x44/0x80
[   75.956190]   do_syscall_64+0x55/0x1a0
[   75.956192]   entry_SYSCALL_64_after_hwframe+0x49/0xbe
[   75.956194] irq event stamp: 2357092
[   75.956196] hardirqs last  enabled at (2357092): [<ffffffff818c62de>] _raw_spin_unlock_irqrestore+0x3e/0x50
[   75.956199] hardirqs last disabled at (2357091): [<ffffffff818c60b1>] _raw_spin_lock_irqsave+0x11/0x80
[   75.956202] softirqs last  enabled at (2357072): [<ffffffff8106dc09>] irq_enter+0x59/0x60
[   75.956204] softirqs last disabled at (2357073): [<ffffffff8106dcbe>] irq_exit+0xae/0xc0
[   75.956206]
               other info that might help us debug this:
[   75.956207]  Possible unsafe locking scenario:
[   75.956208]        CPU0
[   75.956209]        ----
[   75.956210]   lock(&(&da->qlock)->rlock);
[   75.956213]   <Interrupt>
[   75.956214]     lock(&(&da->qlock)->rlock);
[   75.956216]
                *** DEADLOCK ***
[   75.956217] 1 lock held by swapper/0/0:
[   75.956219]  #0: 000000003bb5675c (&(&sc->sc_pcu_lock)->rlock){+.-.}, at: ath9k_tasklet+0x55/0x240 [ath9k]
[   75.956225]
               stack backtrace:
[   75.956228] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.3.0-rc1-wdn+ #13
[   75.956229] Hardware name: Dell Inc. Studio XPS 1340/0K183D, BIOS A11 09/08/2009
[   75.956231] Call Trace:
[   75.956233]  <IRQ>
[   75.956236]  dump_stack+0x67/0x90
[   75.956239]  mark_lock+0x4c1/0x640
[   75.956242]  ? check_usage_backwards+0x130/0x130
[   75.956245]  ? sched_clock_local+0x12/0x80
[   75.956247]  __lock_acquire+0x484/0x7a0
[   75.956250]  ? __lock_acquire+0x3b9/0x7a0
[   75.956252]  lock_acquire+0x9a/0x160
[   75.956259]  ? ath_dynack_sample_ack_ts+0x4d/0xa0 [ath9k_hw]
[   75.956262]  _raw_spin_lock_bh+0x34/0x80
[   75.956268]  ? ath_dynack_sample_ack_ts+0x4d/0xa0 [ath9k_hw]
[   75.956275]  ath_dynack_sample_ack_ts+0x4d/0xa0 [ath9k_hw]
[   75.956280]  ath_rx_tasklet+0xd09/0xe90 [ath9k]
[   75.956286]  ath9k_tasklet+0x102/0x240 [ath9k]
[   75.956288]  tasklet_action_common.isra.0+0x6d/0x170
[   75.956291]  __do_softirq+0xcc/0x425
[   75.956294]  irq_exit+0xae/0xc0
[   75.956296]  do_IRQ+0x8a/0x110
[   75.956298]  common_interrupt+0xf/0xf
[   75.956300]  </IRQ>
[   75.956303] RIP: 0010:cpuidle_enter_state+0xb2/0x400
[   75.956308] RSP: 0018:ffffffff82203e70 EFLAGS: 00000202 ORIG_RAX: ffffffffffffffd7
[   75.956310] RAX: ffffffff82219800 RBX: ffffffff822bd0a0 RCX: 0000000000000000
[   75.956312] RDX: 0000000000000046 RSI: 0000000000000006 RDI: ffffffff82219800
[   75.956314] RBP: ffff888155a01c00 R08: 00000011af51aabe R09: 0000000000000000
[   75.956315] R10: 0000000000000000 R11: 0000000000000000 R12: 0000000000000002
[   75.956317] R13: 00000011af51aabe R14: 0000000000000003 R15: ffffffff82219800
[   75.956321]  cpuidle_enter+0x24/0x40
[   75.956323]  do_idle+0x1ac/0x220
[   75.956326]  cpu_startup_entry+0x14/0x20
[   75.956329]  start_kernel+0x482/0x489
[   75.956332]  secondary_startup_64+0xa4/0xb0
Fixes c774d57fd47c ("ath9k: add dynamic ACK timeout estimation")
---
 drivers/net/wireless/ath/ath9k/dynack.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/net/wireless/ath/ath9k/dynack.c +4 -4
index 6e236a485431..71b4888b30e7 100644
@@ -300,9 +300,9 @@ void ath_dynack_node_init(struct ath_hw *ah, struct ath_node *an)
 
 	an->ackto = ackto;
 
-	spin_lock(&da->qlock);
+	spin_lock_bh(&da->qlock);
 	list_add_tail(&an->list, &da->nodes);
-	spin_unlock(&da->qlock);
+	spin_unlock_bh(&da->qlock);
 }
 EXPORT_SYMBOL(ath_dynack_node_init);
 
@@ -316,9 +316,9 @@ void ath_dynack_node_deinit(struct ath_hw *ah, struct ath_node *an)
 {
 	struct ath_dynack *da = &ah->dynack;
 
-	spin_lock(&da->qlock);
+	spin_lock_bh(&da->qlock);
 	list_del(&an->list);
-	spin_unlock(&da->qlock);
+	spin_unlock_bh(&da->qlock);
 }
 EXPORT_SYMBOL(ath_dynack_node_deinit);
 
Signature · 1 line
2.20.1
author maintainer
#264 lore raw

PATCH 526/671 wcn36xx: use dynamic allocation for large variables

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 355cf31912014e6ff1bb1019ae4858cad12c68cf ]

clang triggers a warning about oversized stack frames that gcc does not notice because of slightly different inlining decisions:

ath/wcn36xx/smd.c:1409:5: error: stack frame size of 1040 bytes in function 'wcn36xx_smd_config_bss' [-Werror,-Wframe-larger-than=] ath/wcn36xx/smd.c:640:5: error: stack frame size of 1032 bytes in function 'wcn36xx_smd_start_hw_scan' [-Werror,-Wframe-larger-than=]

Basically the wcn36xx_hal_start_scan_offload_req_msg, wcn36xx_hal_config_bss_req_msg_v1, and wcn36xx_hal_config_bss_req_msg structures are too large to be put on the kernel stack, but small enough that gcc does not warn about them.

Use kzalloc() to allocate them all. There are similar structures in other parts of this driver, but they are all smaller, with the next largest stack frame at 480 bytes for wcn36xx_smd_send_beacon.

Fixes 8e84c2582169 ("wcn36xx: mac80211 driver for Qualcomm WCN3660/WCN3680 hardware")
---
 drivers/net/wireless/ath/wcn36xx/smd.c | 186 ++++++++++++++-----------
 1 file changed, 105 insertions(+), 81 deletions(-)
drivers/net/wireless/ath/wcn36xx/smd.c +105 -81
index 00098f24116d..6cd113b3b3e4 100644
@@ -641,52 +641,58 @@ int wcn36xx_smd_start_hw_scan(struct wcn36xx *wcn, struct ieee80211_vif *vif,
 			      struct cfg80211_scan_request *req)
 {
 	struct wcn36xx_vif *vif_priv = wcn36xx_vif_to_priv(vif);
-	struct wcn36xx_hal_start_scan_offload_req_msg msg_body;
+	struct wcn36xx_hal_start_scan_offload_req_msg *msg_body;
 	int ret, i;
 
 	if (req->ie_len > WCN36XX_MAX_SCAN_IE_LEN)
 		return -EINVAL;
 
 	mutex_lock(&wcn->hal_mutex);
-	INIT_HAL_MSG(msg_body, WCN36XX_HAL_START_SCAN_OFFLOAD_REQ);
+	msg_body = kzalloc(sizeof(*msg_body), GFP_KERNEL);
+	if (!msg_body) {
+		ret = -ENOMEM;
+		goto out;
+	}
 
-	msg_body.scan_type = WCN36XX_HAL_SCAN_TYPE_ACTIVE;
-	msg_body.min_ch_time = 30;
-	msg_body.max_ch_time = 100;
-	msg_body.scan_hidden = 1;
-	memcpy(msg_body.mac, vif->addr, ETH_ALEN);
-	msg_body.bss_type = vif_priv->bss_type;
-	msg_body.p2p_search = vif->p2p;
+	INIT_HAL_MSG((*msg_body), WCN36XX_HAL_START_SCAN_OFFLOAD_REQ);
 
-	msg_body.num_ssid = min_t(u8, req->n_ssids, ARRAY_SIZE(msg_body.ssids));
-	for (i = 0; i < msg_body.num_ssid; i++) {
-		msg_body.ssids[i].length = min_t(u8, req->ssids[i].ssid_len,
-						sizeof(msg_body.ssids[i].ssid));
-		memcpy(msg_body.ssids[i].ssid, req->ssids[i].ssid,
-		       msg_body.ssids[i].length);
+	msg_body->scan_type = WCN36XX_HAL_SCAN_TYPE_ACTIVE;
+	msg_body->min_ch_time = 30;
+	msg_body->max_ch_time = 100;
+	msg_body->scan_hidden = 1;
+	memcpy(msg_body->mac, vif->addr, ETH_ALEN);
+	msg_body->bss_type = vif_priv->bss_type;
+	msg_body->p2p_search = vif->p2p;
+
+	msg_body->num_ssid = min_t(u8, req->n_ssids, ARRAY_SIZE(msg_body->ssids));
+	for (i = 0; i < msg_body->num_ssid; i++) {
+		msg_body->ssids[i].length = min_t(u8, req->ssids[i].ssid_len,
+						sizeof(msg_body->ssids[i].ssid));
+		memcpy(msg_body->ssids[i].ssid, req->ssids[i].ssid,
+		       msg_body->ssids[i].length);
 	}
 
-	msg_body.num_channel = min_t(u8, req->n_channels,
-				     sizeof(msg_body.channels));
-	for (i = 0; i < msg_body.num_channel; i++)
-		msg_body.channels[i] = req->channels[i]->hw_value;
+	msg_body->num_channel = min_t(u8, req->n_channels,
+				     sizeof(msg_body->channels));
+	for (i = 0; i < msg_body->num_channel; i++)
+		msg_body->channels[i] = req->channels[i]->hw_value;
 
-	msg_body.header.len -= WCN36XX_MAX_SCAN_IE_LEN;
+	msg_body->header.len -= WCN36XX_MAX_SCAN_IE_LEN;
 
 	if (req->ie_len > 0) {
-		msg_body.ie_len = req->ie_len;
-		msg_body.header.len += req->ie_len;
-		memcpy(msg_body.ie, req->ie, req->ie_len);
+		msg_body->ie_len = req->ie_len;
+		msg_body->header.len += req->ie_len;
+		memcpy(msg_body->ie, req->ie, req->ie_len);
 	}
 
-	PREPARE_HAL_BUF(wcn->hal_buf, msg_body);
+	PREPARE_HAL_BUF(wcn->hal_buf, (*msg_body));
 
 	wcn36xx_dbg(WCN36XX_DBG_HAL,
 		    "hal start hw-scan (channels: %u; ssids: %u; p2p: %s)\n",
-		    msg_body.num_channel, msg_body.num_ssid,
-		    msg_body.p2p_search ? "yes" : "no");
+		    msg_body->num_channel, msg_body->num_ssid,
+		    msg_body->p2p_search ? "yes" : "no");
 
-	ret = wcn36xx_smd_send_and_wait(wcn, msg_body.header.len);
+	ret = wcn36xx_smd_send_and_wait(wcn, msg_body->header.len);
 	if (ret) {
 		wcn36xx_err("Sending hal_start_scan_offload failed\n");
 		goto out;
@@ -698,6 +704,7 @@ int wcn36xx_smd_start_hw_scan(struct wcn36xx *wcn, struct ieee80211_vif *vif,
 		goto out;
 	}
 out:
+	kfree(msg_body);
 	mutex_unlock(&wcn->hal_mutex);
 	return ret;
 }
@@ -1257,96 +1264,104 @@ int wcn36xx_smd_config_sta(struct wcn36xx *wcn, struct ieee80211_vif *vif,
 static int wcn36xx_smd_config_bss_v1(struct wcn36xx *wcn,
 			const struct wcn36xx_hal_config_bss_req_msg *orig)
 {
-	struct wcn36xx_hal_config_bss_req_msg_v1 msg_body;
-	struct wcn36xx_hal_config_bss_params_v1 *bss = &msg_body.bss_params;
-	struct wcn36xx_hal_config_sta_params_v1 *sta = &bss->sta;
+	struct wcn36xx_hal_config_bss_req_msg_v1 *msg_body;
+	struct wcn36xx_hal_config_bss_params_v1 *bss;
+	struct wcn36xx_hal_config_sta_params_v1 *sta;
+	int ret;
+
+	msg_body = kzalloc(sizeof(*msg_body), GFP_KERNEL);
+	if (!msg_body)
+		return -ENOMEM;
+
+	INIT_HAL_MSG((*msg_body), WCN36XX_HAL_CONFIG_BSS_REQ);
 
-	INIT_HAL_MSG(msg_body, WCN36XX_HAL_CONFIG_BSS_REQ);
+	bss = &msg_body->bss_params;
+	sta = &bss->sta;
 
 	/* convert orig to v1 */
-	memcpy(&msg_body.bss_params.bssid,
+	memcpy(&msg_body->bss_params.bssid,
 	       &orig->bss_params.bssid, ETH_ALEN);
-	memcpy(&msg_body.bss_params.self_mac_addr,
+	memcpy(&msg_body->bss_params.self_mac_addr,
 	       &orig->bss_params.self_mac_addr, ETH_ALEN);
 
-	msg_body.bss_params.bss_type = orig->bss_params.bss_type;
-	msg_body.bss_params.oper_mode = orig->bss_params.oper_mode;
-	msg_body.bss_params.nw_type = orig->bss_params.nw_type;
+	msg_body->bss_params.bss_type = orig->bss_params.bss_type;
+	msg_body->bss_params.oper_mode = orig->bss_params.oper_mode;
+	msg_body->bss_params.nw_type = orig->bss_params.nw_type;
 
-	msg_body.bss_params.short_slot_time_supported =
+	msg_body->bss_params.short_slot_time_supported =
 		orig->bss_params.short_slot_time_supported;
-	msg_body.bss_params.lla_coexist = orig->bss_params.lla_coexist;
-	msg_body.bss_params.llb_coexist = orig->bss_params.llb_coexist;
-	msg_body.bss_params.llg_coexist = orig->bss_params.llg_coexist;
-	msg_body.bss_params.ht20_coexist = orig->bss_params.ht20_coexist;
-	msg_body.bss_params.lln_non_gf_coexist =
+	msg_body->bss_params.lla_coexist = orig->bss_params.lla_coexist;
+	msg_body->bss_params.llb_coexist = orig->bss_params.llb_coexist;
+	msg_body->bss_params.llg_coexist = orig->bss_params.llg_coexist;
+	msg_body->bss_params.ht20_coexist = orig->bss_params.ht20_coexist;
+	msg_body->bss_params.lln_non_gf_coexist =
 		orig->bss_params.lln_non_gf_coexist;
 
-	msg_body.bss_params.lsig_tx_op_protection_full_support =
+	msg_body->bss_params.lsig_tx_op_protection_full_support =
 		orig->bss_params.lsig_tx_op_protection_full_support;
-	msg_body.bss_params.rifs_mode = orig->bss_params.rifs_mode;
-	msg_body.bss_params.beacon_interval = orig->bss_params.beacon_interval;
-	msg_body.bss_params.dtim_period = orig->bss_params.dtim_period;
-	msg_body.bss_params.tx_channel_width_set =
+	msg_body->bss_params.rifs_mode = orig->bss_params.rifs_mode;
+	msg_body->bss_params.beacon_interval = orig->bss_params.beacon_interval;
+	msg_body->bss_params.dtim_period = orig->bss_params.dtim_period;
+	msg_body->bss_params.tx_channel_width_set =
 		orig->bss_params.tx_channel_width_set;
-	msg_body.bss_params.oper_channel = orig->bss_params.oper_channel;
-	msg_body.bss_params.ext_channel = orig->bss_params.ext_channel;
+	msg_body->bss_params.oper_channel = orig->bss_params.oper_channel;
+	msg_body->bss_params.ext_channel = orig->bss_params.ext_channel;
 
-	msg_body.bss_params.reserved = orig->bss_params.reserved;
+	msg_body->bss_params.reserved = orig->bss_params.reserved;
 
-	memcpy(&msg_body.bss_params.ssid,
+	memcpy(&msg_body->bss_params.ssid,
 	       &orig->bss_params.ssid,
 	       sizeof(orig->bss_params.ssid));
 
-	msg_body.bss_params.action = orig->bss_params.action;
-	msg_body.bss_params.rateset = orig->bss_params.rateset;
-	msg_body.bss_params.ht = orig->bss_params.ht;
-	msg_body.bss_params.obss_prot_enabled =
+	msg_body->bss_params.action = orig->bss_params.action;
+	msg_body->bss_params.rateset = orig->bss_params.rateset;
+	msg_body->bss_params.ht = orig->bss_params.ht;
+	msg_body->bss_params.obss_prot_enabled =
 		orig->bss_params.obss_prot_enabled;
-	msg_body.bss_params.rmf = orig->bss_params.rmf;
-	msg_body.bss_params.ht_oper_mode = orig->bss_params.ht_oper_mode;
-	msg_body.bss_params.dual_cts_protection =
+	msg_body->bss_params.rmf = orig->bss_params.rmf;
+	msg_body->bss_params.ht_oper_mode = orig->bss_params.ht_oper_mode;
+	msg_body->bss_params.dual_cts_protection =
 		orig->bss_params.dual_cts_protection;
 
-	msg_body.bss_params.max_probe_resp_retry_limit =
+	msg_body->bss_params.max_probe_resp_retry_limit =
 		orig->bss_params.max_probe_resp_retry_limit;
-	msg_body.bss_params.hidden_ssid = orig->bss_params.hidden_ssid;
-	msg_body.bss_params.proxy_probe_resp =
+	msg_body->bss_params.hidden_ssid = orig->bss_params.hidden_ssid;
+	msg_body->bss_params.proxy_probe_resp =
 		orig->bss_params.proxy_probe_resp;
-	msg_body.bss_params.edca_params_valid =
+	msg_body->bss_params.edca_params_valid =
 		orig->bss_params.edca_params_valid;
 
-	memcpy(&msg_body.bss_params.acbe,
+	memcpy(&msg_body->bss_params.acbe,
 	       &orig->bss_params.acbe,
 	       sizeof(orig->bss_params.acbe));
-	memcpy(&msg_body.bss_params.acbk,
+	memcpy(&msg_body->bss_params.acbk,
 	       &orig->bss_params.acbk,
 	       sizeof(orig->bss_params.acbk));
-	memcpy(&msg_body.bss_params.acvi,
+	memcpy(&msg_body->bss_params.acvi,
 	       &orig->bss_params.acvi,
 	       sizeof(orig->bss_params.acvi));
-	memcpy(&msg_body.bss_params.acvo,
+	memcpy(&msg_body->bss_params.acvo,
 	       &orig->bss_params.acvo,
 	       sizeof(orig->bss_params.acvo));
 
-	msg_body.bss_params.ext_set_sta_key_param_valid =
+	msg_body->bss_params.ext_set_sta_key_param_valid =
 		orig->bss_params.ext_set_sta_key_param_valid;
 
-	memcpy(&msg_body.bss_params.ext_set_sta_key_param,
+	memcpy(&msg_body->bss_params.ext_set_sta_key_param,
 	       &orig->bss_params.ext_set_sta_key_param,
 	       sizeof(orig->bss_params.acvo));
 
-	msg_body.bss_params.wcn36xx_hal_persona =
+	msg_body->bss_params.wcn36xx_hal_persona =
 		orig->bss_params.wcn36xx_hal_persona;
-	msg_body.bss_params.spectrum_mgt_enable =
+	msg_body->bss_params.spectrum_mgt_enable =
 		orig->bss_params.spectrum_mgt_enable;
-	msg_body.bss_params.tx_mgmt_power = orig->bss_params.tx_mgmt_power;
-	msg_body.bss_params.max_tx_power = orig->bss_params.max_tx_power;
+	msg_body->bss_params.tx_mgmt_power = orig->bss_params.tx_mgmt_power;
+	msg_body->bss_params.max_tx_power = orig->bss_params.max_tx_power;
 
 	wcn36xx_smd_convert_sta_to_v1(wcn, &orig->bss_params.sta,
-				      &msg_body.bss_params.sta);
+				      &msg_body->bss_params.sta);
 
-	PREPARE_HAL_BUF(wcn->hal_buf, msg_body);
+	PREPARE_HAL_BUF(wcn->hal_buf, (*msg_body));
 
 	wcn36xx_dbg(WCN36XX_DBG_HAL,
 		    "hal config bss v1 bssid %pM self_mac_addr %pM bss_type %d oper_mode %d nw_type %d\n",
@@ -1358,7 +1373,10 @@ static int wcn36xx_smd_config_bss_v1(struct wcn36xx *wcn,
 		    sta->bssid, sta->action, sta->sta_index,
 		    sta->bssid_index, sta->aid, sta->type, sta->mac);
 
-	return wcn36xx_smd_send_and_wait(wcn, msg_body.header.len);
+	ret = wcn36xx_smd_send_and_wait(wcn, msg_body->header.len);
+	kfree(msg_body);
+
+	return ret;
 }
 
 
@@ -1410,16 +1428,21 @@ int wcn36xx_smd_config_bss(struct wcn36xx *wcn, struct ieee80211_vif *vif,
 			   struct ieee80211_sta *sta, const u8 *bssid,
 			   bool update)
 {
-	struct wcn36xx_hal_config_bss_req_msg msg;
+	struct wcn36xx_hal_config_bss_req_msg *msg;
 	struct wcn36xx_hal_config_bss_params *bss;
 	struct wcn36xx_hal_config_sta_params *sta_params;
 	struct wcn36xx_vif *vif_priv = wcn36xx_vif_to_priv(vif);
 	int ret;
 
 	mutex_lock(&wcn->hal_mutex);
-	INIT_HAL_MSG(msg, WCN36XX_HAL_CONFIG_BSS_REQ);
+	msg = kzalloc(sizeof(*msg), GFP_KERNEL);
+	if (!msg) {
+		ret = -ENOMEM;
+		goto out;
+	}
+	INIT_HAL_MSG((*msg), WCN36XX_HAL_CONFIG_BSS_REQ);
 
-	bss = &msg.bss_params;
+	bss = &msg->bss_params;
 	sta_params = &bss->sta;
 
 	WARN_ON(is_zero_ether_addr(bssid));
@@ -1514,11 +1537,11 @@ int wcn36xx_smd_config_bss(struct wcn36xx *wcn, struct ieee80211_vif *vif,
 		    sta_params->mac);
 
 	if (!wcn36xx_is_fw_version(wcn, 1, 2, 2, 24)) {
-		ret = wcn36xx_smd_config_bss_v1(wcn, &msg);
+		ret = wcn36xx_smd_config_bss_v1(wcn, msg);
 	} else {
-		PREPARE_HAL_BUF(wcn->hal_buf, msg);
+		PREPARE_HAL_BUF(wcn->hal_buf, (*msg));
 
-		ret = wcn36xx_smd_send_and_wait(wcn, msg.header.len);
+		ret = wcn36xx_smd_send_and_wait(wcn, msg->header.len);
 	}
 	if (ret) {
 		wcn36xx_err("Sending hal_config_bss failed\n");
@@ -1534,6 +1557,7 @@ int wcn36xx_smd_config_bss(struct wcn36xx *wcn, struct ieee80211_vif *vif,
 		goto out;
 	}
 out:
+	kfree(msg);
 	mutex_unlock(&wcn->hal_mutex);
 	return ret;
 }
Signature · 1 line
2.20.1
author maintainer
#265 lore raw

PATCH 527/671 tty: serial: fsl_lpuart: Use appropriate lpuart32_* I/O funcs

In reply to Sasha Levin

From: Andrey Smirnov <[email protected]>

[ Upstream commit 1da17d7cf8e2c4b60163d54300f72c02f510327c ]

When dealing with 32-bit variant of LPUART IP block appropriate I/O helpers have to be used to properly deal with endianness differences. Change all of the offending code to do that.

Fixes a5fa2660d787 ("tty/serial/fsl_lpuart: Add CONSOLE_POLL support

for lpuart32.")

---
 drivers/tty/serial/fsl_lpuart.c | 28 ++++++++++++++--------------
 1 file changed, 14 insertions(+), 14 deletions(-)
drivers/tty/serial/fsl_lpuart.c +14 -14
index ee8a5cb61a5f..2daccb10ae2f 100644
@@ -528,26 +528,26 @@ static int lpuart32_poll_init(struct uart_port *port)
 	spin_lock_irqsave(&sport->port.lock, flags);
 
 	/* Disable Rx & Tx */
-	writel(0, sport->port.membase + UARTCTRL);
+	lpuart32_write(&sport->port, UARTCTRL, 0);
 
-	temp = readl(sport->port.membase + UARTFIFO);
+	temp = lpuart32_read(&sport->port, UARTFIFO);
 
 	/* Enable Rx and Tx FIFO */
-	writel(temp | UARTFIFO_RXFE | UARTFIFO_TXFE,
-		   sport->port.membase + UARTFIFO);
+	lpuart32_write(&sport->port, UARTFIFO,
+		       temp | UARTFIFO_RXFE | UARTFIFO_TXFE);
 
 	/* flush Tx and Rx FIFO */
-	writel(UARTFIFO_TXFLUSH | UARTFIFO_RXFLUSH,
-			sport->port.membase + UARTFIFO);
+	lpuart32_write(&sport->port, UARTFIFO,
+		       UARTFIFO_TXFLUSH | UARTFIFO_RXFLUSH);
 
 	/* explicitly clear RDRF */
-	if (readl(sport->port.membase + UARTSTAT) & UARTSTAT_RDRF) {
-		readl(sport->port.membase + UARTDATA);
-		writel(UARTFIFO_RXUF, sport->port.membase + UARTFIFO);
+	if (lpuart32_read(&sport->port, UARTSTAT) & UARTSTAT_RDRF) {
+		lpuart32_read(&sport->port, UARTDATA);
+		lpuart32_write(&sport->port, UARTFIFO, UARTFIFO_RXUF);
 	}
 
 	/* Enable Rx and Tx */
-	writel(UARTCTRL_RE | UARTCTRL_TE, sport->port.membase + UARTCTRL);
+	lpuart32_write(&sport->port, UARTCTRL, UARTCTRL_RE | UARTCTRL_TE);
 	spin_unlock_irqrestore(&sport->port.lock, flags);
 
 	return 0;
@@ -555,18 +555,18 @@ static int lpuart32_poll_init(struct uart_port *port)
 
 static void lpuart32_poll_put_char(struct uart_port *port, unsigned char c)
 {
-	while (!(readl(port->membase + UARTSTAT) & UARTSTAT_TDRE))
+	while (!(lpuart32_read(port, UARTSTAT) & UARTSTAT_TDRE))
 		barrier();
 
-	writel(c, port->membase + UARTDATA);
+	lpuart32_write(port, UARTDATA, c);
 }
 
 static int lpuart32_poll_get_char(struct uart_port *port)
 {
-	if (!(readl(port->membase + UARTSTAT) & UARTSTAT_RDRF))
+	if (!(lpuart32_read(port, UARTSTAT) & UARTSTAT_RDRF))
 		return NO_POLL_CHAR;
 
-	return readl(port->membase + UARTDATA);
+	return lpuart32_read(port, UARTDATA);
 }
 #endif
 
Signature · 1 line
2.20.1
author maintainer
#266 lore raw

PATCH 528/671 ARM: dts: aspeed-g5: Fixe gpio-ranges upper limit

In reply to Sasha Levin

From: Oscar A Perez <[email protected]>

[ Upstream commit 89b97c429e2e77d695b5133572ca12ec256a4ea4 ]

According to the AST2500/AST2520 specs, these SoCs support up to 228 GPIO pins. However, 'gpio-ranges' value in 'aspeed-g5.dtsi' file is currently setting the upper limit to 220 which isn't allowing access to all their GPIOs. The correct upper limit value is 232 (actual number is 228 plus a 4-GPIO hole in GPIOAB). Without this patch, GPIOs AC5 and AC6 do not work correctly on a AST2500 BMC running Linux Kernel v4.19

Fixes 2039f90d136c ("ARM: dts: aspeed-g5: Add gpio controller to devicetree")
---
 arch/arm/boot/dts/aspeed-g5.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/boot/dts/aspeed-g5.dtsi +1 -1
index d107459fc0f8..f2e1015d75ab 100644
@@ -247,7 +247,7 @@
 				compatible = "aspeed,ast2500-gpio";
 				reg = <0x1e780000 0x1000>;
 				interrupts = <20>;
-				gpio-ranges = <&pinctrl 0 0 220>;
+				gpio-ranges = <&pinctrl 0 0 232>;
 				clocks = <&syscon ASPEED_CLK_APB>;
 				interrupt-controller;
 			};
Signature · 1 line
2.20.1
author maintainer
#267 lore raw

PATCH 529/671 xsk: avoid store-tearing when assigning queues

In reply to Sasha Levin

From: Björn Töpel <[email protected]>

[ Upstream commit 94a997637c5b562fa0ca44fca1d2cd02ec08236f ]

Use WRITE_ONCE when doing the store of tx, rx, fq, and cq, to avoid potential store-tearing. These members are read outside of the control mutex in the mmap implementation.

Fixes 37b076933a8e ("xsk: add missing write- and data-dependency barrier")
---
 net/xdp/xsk.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/xdp/xsk.c +1 -1
index c90854bc3048..b580078f04d1 100644
@@ -320,7 +320,7 @@ static int xsk_init_queue(u32 entries, struct xsk_queue **queue,
 
 	/* Make sure queue is ready before it can be seen by others */
 	smp_wmb();
-	*queue = q;
+	WRITE_ONCE(*queue, q);
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#268 lore raw

PATCH 530/671 xsk: avoid store-tearing when assigning umem

In reply to Sasha Levin

From: Björn Töpel <[email protected]>

[ Upstream commit 9764f4b301c3e7eb3b75eec85b73cad449cdbb0d ]

The umem member of struct xdp_sock is read outside of the control mutex, in the mmap implementation, and needs a WRITE_ONCE to avoid potential store-tearing.

Fixes 423f38329d26 ("xsk: add umem fill queue support and mmap")
---
 net/xdp/xsk.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/xdp/xsk.c +2 -2
index b580078f04d1..72caa4fb13f4 100644
@@ -454,7 +454,7 @@ static int xsk_bind(struct socket *sock, struct sockaddr *addr, int addr_len)
 		}
 
 		xdp_get_umem(umem_xs->umem);
-		xs->umem = umem_xs->umem;
+		WRITE_ONCE(xs->umem, umem_xs->umem);
 		sockfd_put(sock);
 	} else if (!xs->umem || !xdp_umem_validate_queues(xs->umem)) {
 		err = -EINVAL;
@@ -534,7 +534,7 @@ static int xsk_setsockopt(struct socket *sock, int level, int optname,
 
 		/* Make sure umem is ready before it can be seen by others */
 		smp_wmb();
-		xs->umem = umem;
+		WRITE_ONCE(xs->umem, umem);
 		mutex_unlock(&xs->mutex);
 		return 0;
 	}
Signature · 1 line
2.20.1
author maintainer
#269 lore raw

PATCH 531/671 led: triggers: Fix dereferencing of null pointer

In reply to Sasha Levin

From: Oleh Kravchenko <[email protected]>

[ Upstream commit 4016ba85880b252365d11bc7dc899450f2c73ad7 ]

Error was detected by PVS-Studio:
V522 Dereferencing of the null pointer 'led_cdev->trigger' might take place.

Fixes 2282e125a406 ("leds: triggers: let struct led_trigger::activate() return an error code")
---
 drivers/leds/led-triggers.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/leds/led-triggers.c +2 -2
index e4cb3811e82a..005b839f6eb9 100644
@@ -171,11 +171,11 @@ int led_trigger_set(struct led_classdev *led_cdev, struct led_trigger *trig)
 		trig->deactivate(led_cdev);
 err_activate:
 
-	led_cdev->trigger = NULL;
-	led_cdev->trigger_data = NULL;
 	write_lock_irqsave(&led_cdev->trigger->leddev_list_lock, flags);
 	list_del(&led_cdev->trig_list);
 	write_unlock_irqrestore(&led_cdev->trigger->leddev_list_lock, flags);
+	led_cdev->trigger = NULL;
+	led_cdev->trigger_data = NULL;
 	led_set_brightness(led_cdev, LED_OFF);
 	kfree(event);
 
Signature · 1 line
2.20.1
author maintainer
#270 lore raw

PATCH 532/671 net: sonic: return NETDEV_TX_OK if failed to map buffer

In reply to Sasha Levin

From: Mao Wenan <[email protected]>

[ Upstream commit 6e1cdedcf0362fed3aedfe051d46bd7ee2a85fe1 ]

NETDEV_TX_BUSY really should only be used by drivers that call netif_tx_stop_queue() at the wrong moment. If dma_map_single() is failed to map tx DMA buffer, it might trigger an infinite loop. This patch use NETDEV_TX_OK instead of NETDEV_TX_BUSY, and change printk to pr_err_ratelimited.

Fixes d9fb9f384292 ("*sonic/natsemi/ns83829: Move the National Semi-conductor drivers")
---
 drivers/net/ethernet/natsemi/sonic.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/ethernet/natsemi/sonic.c +2 -2
index c805dcbebd02..be36f7117d48 100644
@@ -231,9 +231,9 @@ static int sonic_send_packet(struct sk_buff *skb, struct net_device *dev)
 
 	laddr = dma_map_single(lp->device, skb->data, length, DMA_TO_DEVICE);
 	if (!laddr) {
-		printk(KERN_ERR "%s: failed to map tx DMA buffer.\n", dev->name);
+		pr_err_ratelimited("%s: failed to map tx DMA buffer.\n", dev->name);
 		dev_kfree_skb(skb);
-		return NETDEV_TX_BUSY;
+		return NETDEV_TX_OK;
 	}
 
 	sonic_tda_put(dev, entry, SONIC_TD_STATUS, 0);       /* clear status */
Signature · 1 line
2.20.1
author maintainer
#271 lore raw

PATCH 533/671 net: hns3: fix error VF index when setting VLAN offload

In reply to Sasha Levin

From: Jian Shen <[email protected]>

[ Upstream commit d9c0f2756a33833b2653f7a3612814fa5f52a568 ]

In original codes, the VF index used incorrectly in function hclge_set_vlan_rx_offload_cfg() and hclge_set_vlan_rx_offload_cfg(). When VF id is greater than 8, for example 9, it will set the same bit with VF id 1.

This patch fixes it by using vport->vport_id % HCLGE_VF_NUM_PER_CMD / HCLGE_VF_NUM_PER_BYTE as the array index, instead of vport->vport_id / HCLGE_VF_NUM_PER_CMD.

Fixes 052ece6dc19c ("net: hns3: add ethtool related offload command")
---
 .../ethernet/hisilicon/hns3/hns3pf/hclge_main.c    | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
drivers/net/ethernet/hisilicon/hns3/hns3pf/hclge_main.c +10 -4
index 4b9f898a1620..d575dd9a329d 100644
@@ -4833,6 +4833,7 @@ static int hclge_set_vlan_tx_offload_cfg(struct hclge_vport *vport)
 	struct hclge_vport_vtag_tx_cfg_cmd *req;
 	struct hclge_dev *hdev = vport->back;
 	struct hclge_desc desc;
+	u16 bmap_index;
 	int status;
 
 	hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_VLAN_PORT_TX_CFG, false);
@@ -4855,8 +4856,10 @@ static int hclge_set_vlan_tx_offload_cfg(struct hclge_vport *vport)
 	hnae3_set_bit(req->vport_vlan_cfg, HCLGE_CFG_NIC_ROCE_SEL_B, 0);
 
 	req->vf_offset = vport->vport_id / HCLGE_VF_NUM_PER_CMD;
-	req->vf_bitmap[req->vf_offset] =
-		1 << (vport->vport_id % HCLGE_VF_NUM_PER_BYTE);
+	bmap_index = vport->vport_id % HCLGE_VF_NUM_PER_CMD /
+			HCLGE_VF_NUM_PER_BYTE;
+	req->vf_bitmap[bmap_index] =
+		1U << (vport->vport_id % HCLGE_VF_NUM_PER_BYTE);
 
 	status = hclge_cmd_send(&hdev->hw, &desc, 1);
 	if (status)
@@ -4873,6 +4876,7 @@ static int hclge_set_vlan_rx_offload_cfg(struct hclge_vport *vport)
 	struct hclge_vport_vtag_rx_cfg_cmd *req;
 	struct hclge_dev *hdev = vport->back;
 	struct hclge_desc desc;
+	u16 bmap_index;
 	int status;
 
 	hclge_cmd_setup_basic_desc(&desc, HCLGE_OPC_VLAN_PORT_RX_CFG, false);
@@ -4888,8 +4892,10 @@ static int hclge_set_vlan_rx_offload_cfg(struct hclge_vport *vport)
 		      vcfg->vlan2_vlan_prionly ? 1 : 0);
 
 	req->vf_offset = vport->vport_id / HCLGE_VF_NUM_PER_CMD;
-	req->vf_bitmap[req->vf_offset] =
-		1 << (vport->vport_id % HCLGE_VF_NUM_PER_BYTE);
+	bmap_index = vport->vport_id % HCLGE_VF_NUM_PER_CMD /
+			HCLGE_VF_NUM_PER_BYTE;
+	req->vf_bitmap[bmap_index] =
+		1U << (vport->vport_id % HCLGE_VF_NUM_PER_BYTE);
 
 	status = hclge_cmd_send(&hdev->hw, &desc, 1);
 	if (status)
Signature · 1 line
2.20.1
author maintainer
#272 lore raw

PATCH 534/671 rtlwifi: Fix file release memory leak

In reply to Sasha Levin

From: Wei Yongjun <[email protected]>

[ Upstream commit 4c3e48794dec7cb568974ba3bf2ab62b9c45ca3e ]

When using single_open() for opening, single_release() should be used instead of seq_release(), otherwise there is a memory leak.

This is detected by Coccinelle semantic patch.

Fixes 610247f46feb ("rtlwifi: Improve debugging by using debugfs")
---
 drivers/net/wireless/realtek/rtlwifi/debug.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/realtek/rtlwifi/debug.c +1 -1
index d70385be9976..498994041bbc 100644
@@ -109,7 +109,7 @@ static const struct file_operations file_ops_common = {
 	.open = dl_debug_open_common,
 	.read = seq_read,
 	.llseek = seq_lseek,
-	.release = seq_release,
+	.release = single_release,
 };
 
 static int rtl_debug_get_mac_page(struct seq_file *m, void *v)
Signature · 1 line
2.20.1
author maintainer
#273 lore raw

PATCH 535/671 ARM: dts: logicpd-som-lv: Fix i2c2 and i2c3 Pin mux

In reply to Sasha Levin

From: Adam Ford <[email protected]>

[ Upstream commit a932b77b4d1939ad173f18be87da409427fb705c ]

When the pinmux configuration was added, it was accidentally placed into the omap3_pmx_wkup node when it should have been placed into the omap3_pmx_core. This error was accidentally propagated to stable by me when I blindly requested the pull after seeing I2C issues without actually reviewing the content of the pinout. Since the bootloader previously muxed these correctly in the past, was a hidden error.

This patch moves the i2c2_pins and i2c3_pins to the correct node which should eliminate i2c bus errors and timeouts due to the fact the bootloader uses the save device tree that no longer properly assigns these pins.

Fixes 5fe3c0fa0d54 ("ARM: dts: Add pinmuxing for i2c2 and i2c3

for LogicPD SOM-LV") #4.9+

---
 arch/arm/boot/dts/logicpd-som-lv.dtsi | 26 ++++++++++++++------------
 1 file changed, 14 insertions(+), 12 deletions(-)
arch/arm/boot/dts/logicpd-som-lv.dtsi +14 -12
index 98b682a8080c..c5d54c4d3747 100644
@@ -232,6 +232,20 @@
 		>;
 	};
 
+	i2c2_pins: pinmux_i2c2_pins {
+		pinctrl-single,pins = <
+			OMAP3_CORE1_IOPAD(0x21be, PIN_INPUT | MUX_MODE0)	/* i2c2_scl */
+			OMAP3_CORE1_IOPAD(0x21c0, PIN_INPUT | MUX_MODE0)	/* i2c2_sda */
+		>;
+	};
+
+	i2c3_pins: pinmux_i2c3_pins {
+		pinctrl-single,pins = <
+			OMAP3_CORE1_IOPAD(0x21c2, PIN_INPUT | MUX_MODE0)	/* i2c3_scl */
+			OMAP3_CORE1_IOPAD(0x21c4, PIN_INPUT | MUX_MODE0)	/* i2c3_sda */
+		>;
+	};
+
 	tsc2004_pins: pinmux_tsc2004_pins {
 		pinctrl-single,pins = <
 			OMAP3_CORE1_IOPAD(0x2186, PIN_INPUT | MUX_MODE4)	/* mcbsp4_dr.gpio_153 */
@@ -253,18 +267,6 @@
 			OMAP3_WKUP_IOPAD(0x2a0c, PIN_OUTPUT | MUX_MODE4)	/* sys_boot1.gpio_3 */
 		>;
 	};
-	i2c2_pins: pinmux_i2c2_pins {
-		pinctrl-single,pins = <
-			OMAP3_CORE1_IOPAD(0x21be, PIN_INPUT | MUX_MODE0)	/* i2c2_scl */
-			OMAP3_CORE1_IOPAD(0x21c0, PIN_INPUT | MUX_MODE0)	/* i2c2_sda */
-		>;
-	};
-	i2c3_pins: pinmux_i2c3_pins {
-		pinctrl-single,pins = <
-			OMAP3_CORE1_IOPAD(0x21c2, PIN_INPUT | MUX_MODE0)	/* i2c3_scl */
-			OMAP3_CORE1_IOPAD(0x21c4, PIN_INPUT | MUX_MODE0)	/* i2c3_sda */
-		>;
-	};
 };
 
 &omap3_pmx_core2 {
Signature · 1 line
2.20.1
author maintainer
#274 lore raw

PATCH 536/671 f2fs: fix wrong error injection path in inc_valid_block_count()

In reply to Sasha Levin

From: Chao Yu <[email protected]>

[ Upstream commit 9ea2f0be6ceaebae1518a5f897cff2645830dd95 ]

If FAULT_BLOCK type error injection is on, in inc_valid_block_count() we may decrease sbi->alloc_valid_block_count percpu stat count incorrectly, fix it.

Fixes 36b877af7992 ("f2fs: Keep alloc_valid_block_count in sync")
---
 fs/f2fs/f2fs.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/f2fs/f2fs.h +2 -1
index 72d154e71bb5..6b5b685af599 100644
@@ -1701,7 +1701,7 @@ static inline int inc_valid_block_count(struct f2fs_sb_info *sbi,
 	if (time_to_inject(sbi, FAULT_BLOCK)) {
 		f2fs_show_injection_info(FAULT_BLOCK);
 		release = *count;
-		goto enospc;
+		goto release_quota;
 	}
 
 	/*
@@ -1741,6 +1741,7 @@ static inline int inc_valid_block_count(struct f2fs_sb_info *sbi,
 
 enospc:
 	percpu_counter_sub(&sbi->alloc_valid_block_count, release);
+release_quota:
 	dquot_release_reservation_block(inode, release);
 	return -ENOSPC;
 }
Signature · 8 lines
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
author maintainer
#275 lore raw

PATCH 537/671 f2fs: fix error path of f2fs_convert_inline_page()

In reply to Sasha Levin

From: Chao Yu <[email protected]>

[ Upstream commit e8c82c11c93d586d03d80305959527bcac383555 ]

In error path of f2fs_convert_inline_page(), we missed to truncate newly reserved block in .i_addrs[0] once we failed in get_node_info(), fix it.

Fixes 7735730d39d7 ("f2fs: fix to propagate error from __get_meta_page()")
---
 fs/f2fs/inline.c | 1 +
 1 file changed, 1 insertion(+)
fs/f2fs/inline.c +1 -0
index 6bbb5f6801e2..3fe0dd531390 100644
@@ -133,6 +133,7 @@ int f2fs_convert_inline_page(struct dnode_of_data *dn, struct page *page)
 
 	err = f2fs_get_node_info(fio.sbi, dn->nid, &ni);
 	if (err) {
+		f2fs_truncate_data_blocks_range(dn, 1);
 		f2fs_put_dnode(dn);
 		return err;
 	}
Signature · 1 line
2.20.1
author maintainer
#276 lore raw

PATCH 538/671 scsi: fnic: fix msix interrupt allocation

In reply to Sasha Levin

From: Govindarajulu Varadarajan <[email protected]>

[ Upstream commit 3ec24fb4c035e9cbb2f02a48640a09aa913442a2 ]

pci_alloc_irq_vectors() returns number of vectors allocated. Fix the check for error condition.

---
 drivers/scsi/fnic/fnic_isr.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/scsi/fnic/fnic_isr.c +2 -2
index 4e3a50202e8c..d28088218c36 100644
@@ -254,7 +254,7 @@ int fnic_set_intr_mode(struct fnic *fnic)
 		int vecs = n + m + o + 1;
 
 		if (pci_alloc_irq_vectors(fnic->pdev, vecs, vecs,
-				PCI_IRQ_MSIX) < 0) {
+				PCI_IRQ_MSIX) == vecs) {
 			fnic->rq_count = n;
 			fnic->raw_wq_count = m;
 			fnic->wq_copy_count = o;
@@ -280,7 +280,7 @@ int fnic_set_intr_mode(struct fnic *fnic)
 	    fnic->wq_copy_count >= 1 &&
 	    fnic->cq_count >= 3 &&
 	    fnic->intr_count >= 1 &&
-	    pci_alloc_irq_vectors(fnic->pdev, 1, 1, PCI_IRQ_MSI) < 0) {
+	    pci_alloc_irq_vectors(fnic->pdev, 1, 1, PCI_IRQ_MSI) == 1) {
 		fnic->rq_count = 1;
 		fnic->raw_wq_count = 1;
 		fnic->wq_copy_count = 1;
Signature · 1 line
2.20.1
author maintainer
#277 lore raw

PATCH 539/671 Btrfs: fix hang when loading existing inode cache off disk

In reply to Sasha Levin

From: Filipe Manana <[email protected]>

[ Upstream commit 7764d56baa844d7f6206394f21a0e8c1f303c476 ]

If we are able to load an existing inode cache off disk, we set the state of the cache to BTRFS_CACHE_FINISHED, but we don't wake up any one waiting for the cache to be available. This means that anyone waiting for the cache to be available, waiting on the condition that either its state is BTRFS_CACHE_FINISHED or its available free space is greather than zero, can hang forever.

This could be observed running fstests with MOUNT_OPTIONS="-o inode_cache", in particular test case generic/161 triggered it very frequently for me, producing a trace like the following:

  [63795.739712] BTRFS info (device sdc): enabling inode map caching
  [63795.739714] BTRFS info (device sdc): disk space caching is enabled
  [63795.739716] BTRFS info (device sdc): has skinny extents
  [64036.653886] INFO: task btrfs-transacti:3917 blocked for more than 120 seconds.
  [64036.654079]       Not tainted 5.2.0-rc4-btrfs-next-50 #1
  [64036.654143] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
  [64036.654232] btrfs-transacti D    0  3917      2 0x80004000
  [64036.654239] Call Trace:
  [64036.654258]  ? __schedule+0x3ae/0x7b0
  [64036.654271]  schedule+0x3a/0xb0
  [64036.654325]  btrfs_commit_transaction+0x978/0xae0 [btrfs]
  [64036.654339]  ? remove_wait_queue+0x60/0x60
  [64036.654395]  transaction_kthread+0x146/0x180 [btrfs]
  [64036.654450]  ? btrfs_cleanup_transaction+0x620/0x620 [btrfs]
  [64036.654456]  kthread+0x103/0x140
  [64036.654464]  ? kthread_create_worker_on_cpu+0x70/0x70
  [64036.654476]  ret_from_fork+0x3a/0x50
  [64036.654504] INFO: task xfs_io:3919 blocked for more than 120 seconds.
  [64036.654568]       Not tainted 5.2.0-rc4-btrfs-next-50 #1
  [64036.654617] "echo 0 > /proc/sys/kernel/hung_task_timeout_secs" disables this message.
  [64036.654685] xfs_io          D    0  3919   3633 0x00000000
  [64036.654691] Call Trace:
  [64036.654703]  ? __schedule+0x3ae/0x7b0
  [64036.654716]  schedule+0x3a/0xb0
  [64036.654756]  btrfs_find_free_ino+0xa9/0x120 [btrfs]
  [64036.654764]  ? remove_wait_queue+0x60/0x60
  [64036.654809]  btrfs_create+0x72/0x1f0 [btrfs]
  [64036.654822]  lookup_open+0x6bc/0x790
  [64036.654849]  path_openat+0x3bc/0xc00
  [64036.654854]  ? __lock_acquire+0x331/0x1cb0
  [64036.654869]  do_filp_open+0x99/0x110
  [64036.654884]  ? __alloc_fd+0xee/0x200
  [64036.654895]  ? do_raw_spin_unlock+0x49/0xc0
  [64036.654909]  ? do_sys_open+0x132/0x220
  [64036.654913]  do_sys_open+0x132/0x220
  [64036.654926]  do_syscall_64+0x60/0x1d0
  [64036.654933]  entry_SYSCALL_64_after_hwframe+0x49/0xbe

Fix this by adding a wake_up() call right after setting the cache state to BTRFS_CACHE_FINISHED, at start_caching(), when we are able to load the cache from disk.

Fixes 82d5902d9c681b ("Btrfs: Support reading/writing on disk free ino cache")
---
 fs/btrfs/inode-map.c | 1 +
 1 file changed, 1 insertion(+)
fs/btrfs/inode-map.c +1 -0
index e1b50c62ba65..b3bd27070617 100644
@@ -145,6 +145,7 @@ static void start_caching(struct btrfs_root *root)
 		spin_lock(&root->ino_cache_lock);
 		root->ino_cache_state = BTRFS_CACHE_FINISHED;
 		spin_unlock(&root->ino_cache_lock);
+		wake_up(&root->ino_cache_wait);
 		return;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#278 lore raw

PATCH 540/671 Btrfs: fix inode cache waiters hanging on failure to start caching thread

In reply to Sasha Levin

From: Filipe Manana <[email protected]>

[ Upstream commit a68ebe0790fc88b4314d17984a2cf99ce2361901 ]

If we fail to start the inode caching thread, we print an error message and disable the inode cache, however we never wake up any waiters, so they hang forever waiting for the caching to finish. Fix this by waking them up and have them fallback to a call to btrfs_find_free_objectid().

Fixes e60efa84252c05 ("Btrfs: avoid triggering bug_on() when we fail to start inode caching task")
---
 fs/btrfs/inode-map.c | 23 ++++++++++++++++++-----
 1 file changed, 18 insertions(+), 5 deletions(-)
fs/btrfs/inode-map.c +18 -5
index b3bd27070617..7c4d0107c6fb 100644
@@ -12,6 +12,19 @@
 #include "inode-map.h"
 #include "transaction.h"
 
+static void fail_caching_thread(struct btrfs_root *root)
+{
+	struct btrfs_fs_info *fs_info = root->fs_info;
+
+	btrfs_warn(fs_info, "failed to start inode caching task");
+	btrfs_clear_pending_and_info(fs_info, INODE_MAP_CACHE,
+				     "disabling inode map caching");
+	spin_lock(&root->ino_cache_lock);
+	root->ino_cache_state = BTRFS_CACHE_ERROR;
+	spin_unlock(&root->ino_cache_lock);
+	wake_up(&root->ino_cache_wait);
+}
+
 static int caching_kthread(void *data)
 {
 	struct btrfs_root *root = data;
@@ -164,11 +177,8 @@ static void start_caching(struct btrfs_root *root)
 
 	tsk = kthread_run(caching_kthread, root, "btrfs-ino-cache-%llu",
 			  root->root_key.objectid);
-	if (IS_ERR(tsk)) {
-		btrfs_warn(fs_info, "failed to start inode caching task");
-		btrfs_clear_pending_and_info(fs_info, INODE_MAP_CACHE,
-					     "disabling inode map caching");
-	}
+	if (IS_ERR(tsk))
+		fail_caching_thread(root);
 }
 
 int btrfs_find_free_ino(struct btrfs_root *root, u64 *objectid)
@@ -186,11 +196,14 @@ int btrfs_find_free_ino(struct btrfs_root *root, u64 *objectid)
 
 	wait_event(root->ino_cache_wait,
 		   root->ino_cache_state == BTRFS_CACHE_FINISHED ||
+		   root->ino_cache_state == BTRFS_CACHE_ERROR ||
 		   root->free_ino_ctl->free_space > 0);
 
 	if (root->ino_cache_state == BTRFS_CACHE_FINISHED &&
 	    root->free_ino_ctl->free_space == 0)
 		return -ENOSPC;
+	else if (root->ino_cache_state == BTRFS_CACHE_ERROR)
+		return btrfs_find_free_objectid(root, objectid);
 	else
 		goto again;
 }
Signature · 1 line
2.20.1
author maintainer
#279 lore raw

PATCH 541/671 Btrfs: fix inode cache waiters hanging on path allocation failure

In reply to Sasha Levin

From: Filipe Manana <[email protected]>

[ Upstream commit 9d123a35d7e97bb2139747b16127c9b22b6a593e ]

If the caching thread fails to allocate a path, it returns without waking up any cache waiters, leaving them hang forever. Fix this by following the same approach as when we fail to start the caching thread: print an error message, disable inode caching and make the wakers fallback to non-caching mode behaviour (calling btrfs_find_free_objectid()).

Fixes 581bb050941b4f ("Btrfs: Cache free inode numbers in memory")
---
 fs/btrfs/inode-map.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
fs/btrfs/inode-map.c +3 -1
index 7c4d0107c6fb..c7dcb7c52108 100644
@@ -41,8 +41,10 @@ static int caching_kthread(void *data)
 		return 0;
 
 	path = btrfs_alloc_path();
-	if (!path)
+	if (!path) {
+		fail_caching_thread(root);
 		return -ENOMEM;
+	}
 
 	/* Since the commit root is read-only, we can safely skip locking. */
 	path->skip_locking = 1;
Signature · 1 line
2.20.1
author maintainer
#280 lore raw

PATCH 542/671 btrfs: use correct count in btrfs_file_write_iter()

In reply to Sasha Levin

From: Omar Sandoval <[email protected]>

[ Upstream commit c09767a8960ca0500fb636bf73686723337debf4 ]

generic_write_checks() may modify iov_iter_count(), so we must get the count after the call, not before. Using the wrong one has a couple of consequences:

1. We check a longer range in check_can_nocow() for nowait than we're actually writing.
2. We create extra hole extent maps in btrfs_cont_expand(). As far as I can tell, this is harmless, but I might be missing something.

These issues are pretty minor, but let's fix it before something more important trips on it.

Fixes edf064e7c6fe ("btrfs: nowait aio support")
---
 fs/btrfs/file.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
fs/btrfs/file.c +2 -1
index d9d90f0b66d2..f4a7dac76356 100644
@@ -1895,7 +1895,7 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb,
 	bool sync = (file->f_flags & O_DSYNC) || IS_SYNC(file->f_mapping->host);
 	ssize_t err;
 	loff_t pos;
-	size_t count = iov_iter_count(from);
+	size_t count;
 	loff_t oldsize;
 	int clean_page = 0;
 
@@ -1916,6 +1916,7 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb,
 	}
 
 	pos = iocb->ki_pos;
+	count = iov_iter_count(from);
 	if (iocb->ki_flags & IOCB_NOWAIT) {
 		/*
 		 * We will allocate space in case nodatacow is not set,
Signature · 1 line
2.20.1
author maintainer
#281 lore raw

PATCH 543/671 ixgbe: sync the first fragment unconditionally

In reply to Sasha Levin

From: Firo Yang <[email protected]>

[ Upstream commit e7ba676c6188d394a0133fc4b9bcd7ee50d54b7f ]

In Xen environment, if Xen-swiotlb is enabled, ixgbe driver could possibly allocate a page, DMA memory buffer, for the first fragment which is not suitable for Xen-swiotlb to do DMA operations. Xen-swiotlb have to internally allocate another page for doing DMA operations. This mechanism requires syncing the data from the internal page to the page which ixgbe sends to upper network stack. However, since commit f3213d932173 ("ixgbe: Update driver to make use of DMA attributes in Rx path"), the unmap operation is performed with DMA_ATTR_SKIP_CPU_SYNC. As a result, the sync is not performed. Since the sync isn't performed, the upper network stack could receive a incomplete network packet. By incomplete, it means the linear data on the first fragment(between skb->head and skb->end) is invalid. So we have to copy the data from the internal xen-swiotlb page to the page which ixgbe sends to upper network stack through the sync operation.

More details from Alexander Duyck:
Specifically since we are mapping the frame with DMA_ATTR_SKIP_CPU_SYNC we have to unmap with that as well. As a result a sync is not performed on an unmap and must be done manually as we skipped it for the first frag. As such we need to always sync before possibly performing a page unmap operation.

Fixes f3213d932173 ("ixgbe: Update driver to make use of DMA attributes in Rx path")
---
 drivers/net/ethernet/intel/ixgbe/ixgbe_main.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
drivers/net/ethernet/intel/ixgbe/ixgbe_main.c +9 -7
index de65ca1e6558..51cd58fbab69 100644
@@ -1822,13 +1822,7 @@ static void ixgbe_pull_tail(struct ixgbe_ring *rx_ring,
 static void ixgbe_dma_sync_frag(struct ixgbe_ring *rx_ring,
 				struct sk_buff *skb)
 {
-	/* if the page was released unmap it, else just sync our portion */
-	if (unlikely(IXGBE_CB(skb)->page_released)) {
-		dma_unmap_page_attrs(rx_ring->dev, IXGBE_CB(skb)->dma,
-				     ixgbe_rx_pg_size(rx_ring),
-				     DMA_FROM_DEVICE,
-				     IXGBE_RX_DMA_ATTR);
-	} else if (ring_uses_build_skb(rx_ring)) {
+	if (ring_uses_build_skb(rx_ring)) {
 		unsigned long offset = (unsigned long)(skb->data) & ~PAGE_MASK;
 
 		dma_sync_single_range_for_cpu(rx_ring->dev,
@@ -1845,6 +1839,14 @@ static void ixgbe_dma_sync_frag(struct ixgbe_ring *rx_ring,
 					      skb_frag_size(frag),
 					      DMA_FROM_DEVICE);
 	}
+
+	/* If the page was released, just unmap it. */
+	if (unlikely(IXGBE_CB(skb)->page_released)) {
+		dma_unmap_page_attrs(rx_ring->dev, IXGBE_CB(skb)->dma,
+				     ixgbe_rx_pg_size(rx_ring),
+				     DMA_FROM_DEVICE,
+				     IXGBE_RX_DMA_ATTR);
+	}
 }
 
 /**
Signature · 1 line
2.20.1
author maintainer
#282 lore raw

PATCH 544/671 hwmon: (shtc1) fix shtc1 and shtw1 id mask

In reply to Sasha Levin

From: Dan Robertson <[email protected]>

[ Upstream commit fdc7d8e829ec755c5cfb2f5a8d8c0cdfb664f895 ]

Fix an error in the bitmaskfor the shtc1 and shtw1 bitmask used to retrieve the chip ID from the ID register. See section 5.7 of the shtw1 or shtc1 datasheet for details.

Fixes 1a539d372edd9832444e7a3daa710c444c014dc9 ("hwmon: add support for Sensirion SHTC1 sensor")
[groeck: Reordered to be first in series and adjusted accordingly]
---
 drivers/hwmon/shtc1.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hwmon/shtc1.c +1 -1
index decd7df995ab..2a18539591ea 100644
@@ -38,7 +38,7 @@ static const unsigned char shtc1_cmd_read_id_reg[]	       = { 0xef, 0xc8 };
 
 /* constants for reading the ID register */
 #define SHTC1_ID	  0x07
-#define SHTC1_ID_REG_MASK 0x1f
+#define SHTC1_ID_REG_MASK 0x3f
 
 /* delays for non-blocking i2c commands, both in us */
 #define SHTC1_NONBLOCKING_WAIT_TIME_HPM  14400
Signature · 1 line
2.20.1
author maintainer
#283 lore raw

PATCH 545/671 net: sonic: replace dev_kfree_skb in sonic_send_packet

In reply to Sasha Levin

From: Mao Wenan <[email protected]>

[ Upstream commit 49f6c90bf6805948b597eabb499e500a47cf24be ]

sonic_send_packet will be processed in irq or non-irq context, so it would better use dev_kfree_skb_any instead of dev_kfree_skb.

Fixes d9fb9f384292 ("*sonic/natsemi/ns83829: Move the National Semi-conductor drivers")
---
 drivers/net/ethernet/natsemi/sonic.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/natsemi/sonic.c +1 -1
index be36f7117d48..5f1875fe47cd 100644
@@ -232,7 +232,7 @@ static int sonic_send_packet(struct sk_buff *skb, struct net_device *dev)
 	laddr = dma_map_single(lp->device, skb->data, length, DMA_TO_DEVICE);
 	if (!laddr) {
 		pr_err_ratelimited("%s: failed to map tx DMA buffer.\n", dev->name);
-		dev_kfree_skb(skb);
+		dev_kfree_skb_any(skb);
 		return NETDEV_TX_OK;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#284 lore raw

PATCH 546/671 pinctrl: iproc-gpio: Fix incorrect pinconf configurations

In reply to Sasha Levin

From: Li Jin <[email protected]>

[ Upstream commit 398a1f50e3c731586182fd52b834103b0aa2f826 ]

Fix drive strength for AON/CRMU controller; fix pull-up/down setting for CCM/CDRU controller.

Fixes 616043d58a89 ("pinctrl: Rename gpio driver from cygnus to iproc")
---
 drivers/pinctrl/bcm/pinctrl-iproc-gpio.c | 96 +++++++++++++++++++-----
 1 file changed, 77 insertions(+), 19 deletions(-)
drivers/pinctrl/bcm/pinctrl-iproc-gpio.c +77 -19
index b70058caee50..20b9864adce0 100644
@@ -54,8 +54,12 @@
 /* drive strength control for ASIU GPIO */
 #define IPROC_GPIO_ASIU_DRV0_CTRL_OFFSET 0x58
 
-/* drive strength control for CCM/CRMU (AON) GPIO */
-#define IPROC_GPIO_DRV0_CTRL_OFFSET  0x00
+/* pinconf for CCM GPIO */
+#define IPROC_GPIO_PULL_DN_OFFSET   0x10
+#define IPROC_GPIO_PULL_UP_OFFSET   0x14
+
+/* pinconf for CRMU(aon) GPIO and CCM GPIO*/
+#define IPROC_GPIO_DRV_CTRL_OFFSET  0x00
 
 #define GPIO_BANK_SIZE 0x200
 #define NGPIOS_PER_BANK 32
@@ -76,6 +80,12 @@ enum iproc_pinconf_param {
 	IPROC_PINCON_MAX,
 };
 
+enum iproc_pinconf_ctrl_type {
+	IOCTRL_TYPE_AON = 1,
+	IOCTRL_TYPE_CDRU,
+	IOCTRL_TYPE_INVALID,
+};
+
 /*
  * Iproc GPIO core
  *
@@ -100,6 +110,7 @@ struct iproc_gpio {
 
 	void __iomem *base;
 	void __iomem *io_ctrl;
+	enum iproc_pinconf_ctrl_type io_ctrl_type;
 
 	raw_spinlock_t lock;
 
@@ -461,20 +472,44 @@ static const struct pinctrl_ops iproc_pctrl_ops = {
 static int iproc_gpio_set_pull(struct iproc_gpio *chip, unsigned gpio,
 				bool disable, bool pull_up)
 {
+	void __iomem *base;
 	unsigned long flags;
+	unsigned int shift;
+	u32 val_1, val_2;
 
 	raw_spin_lock_irqsave(&chip->lock, flags);
-
-	if (disable) {
-		iproc_set_bit(chip, IPROC_GPIO_RES_EN_OFFSET, gpio, false);
+	if (chip->io_ctrl_type == IOCTRL_TYPE_CDRU) {
+		base = chip->io_ctrl;
+		shift = IPROC_GPIO_SHIFT(gpio);
+
+		val_1 = readl(base + IPROC_GPIO_PULL_UP_OFFSET);
+		val_2 = readl(base + IPROC_GPIO_PULL_DN_OFFSET);
+		if (disable) {
+			/* no pull-up or pull-down */
+			val_1 &= ~BIT(shift);
+			val_2 &= ~BIT(shift);
+		} else if (pull_up) {
+			val_1 |= BIT(shift);
+			val_2 &= ~BIT(shift);
+		} else {
+			val_1 &= ~BIT(shift);
+			val_2 |= BIT(shift);
+		}
+		writel(val_1, base + IPROC_GPIO_PULL_UP_OFFSET);
+		writel(val_2, base + IPROC_GPIO_PULL_DN_OFFSET);
 	} else {
-		iproc_set_bit(chip, IPROC_GPIO_PAD_RES_OFFSET, gpio,
-			       pull_up);
-		iproc_set_bit(chip, IPROC_GPIO_RES_EN_OFFSET, gpio, true);
+		if (disable) {
+			iproc_set_bit(chip, IPROC_GPIO_RES_EN_OFFSET, gpio,
+				      false);
+		} else {
+			iproc_set_bit(chip, IPROC_GPIO_PAD_RES_OFFSET, gpio,
+				      pull_up);
+			iproc_set_bit(chip, IPROC_GPIO_RES_EN_OFFSET, gpio,
+				      true);
+		}
 	}
 
 	raw_spin_unlock_irqrestore(&chip->lock, flags);
-
 	dev_dbg(chip->dev, "gpio:%u set pullup:%d\n", gpio, pull_up);
 
 	return 0;
@@ -483,14 +518,35 @@ static int iproc_gpio_set_pull(struct iproc_gpio *chip, unsigned gpio,
 static void iproc_gpio_get_pull(struct iproc_gpio *chip, unsigned gpio,
 				 bool *disable, bool *pull_up)
 {
+	void __iomem *base;
 	unsigned long flags;
+	unsigned int shift;
+	u32 val_1, val_2;
 
 	raw_spin_lock_irqsave(&chip->lock, flags);
-	*disable = !iproc_get_bit(chip, IPROC_GPIO_RES_EN_OFFSET, gpio);
-	*pull_up = iproc_get_bit(chip, IPROC_GPIO_PAD_RES_OFFSET, gpio);
+	if (chip->io_ctrl_type == IOCTRL_TYPE_CDRU) {
+		base = chip->io_ctrl;
+		shift = IPROC_GPIO_SHIFT(gpio);
+
+		val_1 = readl(base + IPROC_GPIO_PULL_UP_OFFSET) & BIT(shift);
+		val_2 = readl(base + IPROC_GPIO_PULL_DN_OFFSET) & BIT(shift);
+
+		*pull_up = val_1 ? true : false;
+		*disable = (val_1 | val_2) ? false : true;
+
+	} else {
+		*disable = !iproc_get_bit(chip, IPROC_GPIO_RES_EN_OFFSET, gpio);
+		*pull_up = iproc_get_bit(chip, IPROC_GPIO_PAD_RES_OFFSET, gpio);
+	}
 	raw_spin_unlock_irqrestore(&chip->lock, flags);
 }
 
+#define DRV_STRENGTH_OFFSET(gpio, bit, type)  ((type) == IOCTRL_TYPE_AON ? \
+	((2 - (bit)) * 4 + IPROC_GPIO_DRV_CTRL_OFFSET) : \
+	((type) == IOCTRL_TYPE_CDRU) ? \
+	((bit) * 4 + IPROC_GPIO_DRV_CTRL_OFFSET) : \
+	((bit) * 4 + IPROC_GPIO_REG(gpio, IPROC_GPIO_ASIU_DRV0_CTRL_OFFSET)))
+
 static int iproc_gpio_set_strength(struct iproc_gpio *chip, unsigned gpio,
 				    unsigned strength)
 {
@@ -505,11 +561,8 @@ static int iproc_gpio_set_strength(struct iproc_gpio *chip, unsigned gpio,
 
 	if (chip->io_ctrl) {
 		base = chip->io_ctrl;
-		offset = IPROC_GPIO_DRV0_CTRL_OFFSET;
 	} else {
 		base = chip->base;
-		offset = IPROC_GPIO_REG(gpio,
-					 IPROC_GPIO_ASIU_DRV0_CTRL_OFFSET);
 	}
 
 	shift = IPROC_GPIO_SHIFT(gpio);
@@ -520,11 +573,11 @@ static int iproc_gpio_set_strength(struct iproc_gpio *chip, unsigned gpio,
 	raw_spin_lock_irqsave(&chip->lock, flags);
 	strength = (strength / 2) - 1;
 	for (i = 0; i < GPIO_DRV_STRENGTH_BITS; i++) {
+		offset = DRV_STRENGTH_OFFSET(gpio, i, chip->io_ctrl_type);
 		val = readl(base + offset);
 		val &= ~BIT(shift);
 		val |= ((strength >> i) & 0x1) << shift;
 		writel(val, base + offset);
-		offset += 4;
 	}
 	raw_spin_unlock_irqrestore(&chip->lock, flags);
 
@@ -541,11 +594,8 @@ static int iproc_gpio_get_strength(struct iproc_gpio *chip, unsigned gpio,
 
 	if (chip->io_ctrl) {
 		base = chip->io_ctrl;
-		offset = IPROC_GPIO_DRV0_CTRL_OFFSET;
 	} else {
 		base = chip->base;
-		offset = IPROC_GPIO_REG(gpio,
-					 IPROC_GPIO_ASIU_DRV0_CTRL_OFFSET);
 	}
 
 	shift = IPROC_GPIO_SHIFT(gpio);
@@ -553,10 +603,10 @@ static int iproc_gpio_get_strength(struct iproc_gpio *chip, unsigned gpio,
 	raw_spin_lock_irqsave(&chip->lock, flags);
 	*strength = 0;
 	for (i = 0; i < GPIO_DRV_STRENGTH_BITS; i++) {
+		offset = DRV_STRENGTH_OFFSET(gpio, i, chip->io_ctrl_type);
 		val = readl(base + offset) & BIT(shift);
 		val >>= shift;
 		*strength += (val << i);
-		offset += 4;
 	}
 
 	/* convert to mA */
@@ -734,6 +784,7 @@ static int iproc_gpio_probe(struct platform_device *pdev)
 	u32 ngpios, pinconf_disable_mask = 0;
 	int irq, ret;
 	bool no_pinconf = false;
+	enum iproc_pinconf_ctrl_type io_ctrl_type = IOCTRL_TYPE_INVALID;
 
 	/* NSP does not support drive strength config */
 	if (of_device_is_compatible(dev->of_node, "brcm,iproc-nsp-gpio"))
@@ -764,8 +815,15 @@ static int iproc_gpio_probe(struct platform_device *pdev)
 			dev_err(dev, "unable to map I/O memory\n");
 			return PTR_ERR(chip->io_ctrl);
 		}
+		if (of_device_is_compatible(dev->of_node,
+					    "brcm,cygnus-ccm-gpio"))
+			io_ctrl_type = IOCTRL_TYPE_CDRU;
+		else
+			io_ctrl_type = IOCTRL_TYPE_AON;
 	}
 
+	chip->io_ctrl_type = io_ctrl_type;
+
 	if (of_property_read_u32(dev->of_node, "ngpios", &ngpios)) {
 		dev_err(&pdev->dev, "missing ngpios DT property\n");
 		return -ENODEV;
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#285 lore raw

PATCH 547/671 gpio/aspeed: Fix incorrect number of banks

In reply to Sasha Levin

From: Rashmica Gupta <[email protected]>

[ Upstream commit 3c4710ae6f883f9c6e3df5e27e274702a1221c57 ]

The current calculation for the number of GPIO banks is only correct if the number of GPIOs is a multiple of 32 (if there were 31 GPIOs we would currently say there are 0 banks, which is incorrect).

Fixes 361b79119a4b7 ('gpio: Add Aspeed driver')
---
 drivers/gpio/gpio-aspeed.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpio/gpio-aspeed.c +1 -1
index b696ec35efb3..e627e0e9001a 100644
@@ -1199,7 +1199,7 @@ static int __init aspeed_gpio_probe(struct platform_device *pdev)
 	gpio->chip.irq.need_valid_mask = true;
 
 	/* Allocate a cache of the output registers */
-	banks = gpio->config->nr_gpios >> 5;
+	banks = DIV_ROUND_UP(gpio->config->nr_gpios, 32);
 	gpio->dcache = devm_kcalloc(&pdev->dev,
 				    banks, sizeof(u32), GFP_KERNEL);
 	if (!gpio->dcache)
Signature · 1 line
2.20.1
author maintainer
#286 lore raw

PATCH 548/671 ath10k: adjust skb length in ath10k_sdio_mbox_rx_packet

In reply to Sasha Levin

From: Nicolas Boichat <[email protected]>

[ Upstream commit b7139960832eb56fa15d390a4b5c8c5739bd0d1a ]

When the FW bundles multiple packets, pkt->act_len may be incorrect as it refers to the first packet only (however, the FW will only bundle packets that fit into the same pkt->alloc_len).

Before this patch, the skb length would be set (incorrectly) to pkt->act_len in ath10k_sdio_mbox_rx_packet, and then later manually adjusted in ath10k_sdio_mbox_rx_process_packet.

The first problem is that ath10k_sdio_mbox_rx_process_packet does not use proper skb_put commands to adjust the length (it directly changes skb->len), so we end up with a mismatch between skb->head + skb->tail and skb->data + skb->len. This is quite serious, and causes corruptions in the TCP stack, as the stack tries to coalesce packets, and relies on skb->tail being correct (that is, skb_tail_pointer must point to the first byte_after_ the data).

Instead of re-adjusting the size in ath10k_sdio_mbox_rx_process_packet, this moves the code to ath10k_sdio_mbox_rx_packet, and also add a bounds check, as skb_put would crash the kernel if not enough space is available.

Tested with QCA6174 SDIO with firmware
WLAN.RMH.4.4.1-00007-QCARMSWP-1.

Fixes 8530b4e7b22bc3b ("ath10k: sdio: set skb len for all rx packets")
---
 drivers/net/wireless/ath/ath10k/sdio.c | 29 +++++++++++++++++++-------
 1 file changed, 21 insertions(+), 8 deletions(-)
drivers/net/wireless/ath/ath10k/sdio.c +21 -8
index 686759b5613f..0ecaba824fb2 100644
@@ -392,16 +392,11 @@ static int ath10k_sdio_mbox_rx_process_packet(struct ath10k *ar,
 	struct ath10k_htc_hdr *htc_hdr = (struct ath10k_htc_hdr *)skb->data;
 	bool trailer_present = htc_hdr->flags & ATH10K_HTC_FLAG_TRAILER_PRESENT;
 	enum ath10k_htc_ep_id eid;
-	u16 payload_len;
 	u8 *trailer;
 	int ret;
 
-	payload_len = le16_to_cpu(htc_hdr->len);
-	skb->len = payload_len + sizeof(struct ath10k_htc_hdr);
-
 	if (trailer_present) {
-		trailer = skb->data + sizeof(*htc_hdr) +
-			  payload_len - htc_hdr->trailer_len;
+		trailer = skb->data + skb->len - htc_hdr->trailer_len;
 
 		eid = pipe_id_to_eid(htc_hdr->eid);
 
@@ -638,13 +633,31 @@ static int ath10k_sdio_mbox_rx_packet(struct ath10k *ar,
 {
 	struct ath10k_sdio *ar_sdio = ath10k_sdio_priv(ar);
 	struct sk_buff *skb = pkt->skb;
+	struct ath10k_htc_hdr *htc_hdr;
 	int ret;
 
 	ret = ath10k_sdio_readsb(ar, ar_sdio->mbox_info.htc_addr,
 				 skb->data, pkt->alloc_len);
+	if (ret)
+		goto out;
+
+	/* Update actual length. The original length may be incorrect,
+	 * as the FW will bundle multiple packets as long as their sizes
+	 * fit within the same aligned length (pkt->alloc_len).
+	 */
+	htc_hdr = (struct ath10k_htc_hdr *)skb->data;
+	pkt->act_len = le16_to_cpu(htc_hdr->len) + sizeof(*htc_hdr);
+	if (pkt->act_len > pkt->alloc_len) {
+		ath10k_warn(ar, "rx packet too large (%zu > %zu)\n",
+			    pkt->act_len, pkt->alloc_len);
+		ret = -EMSGSIZE;
+		goto out;
+	}
+
+	skb_put(skb, pkt->act_len);
+
+out:
 	pkt->status = ret;
-	if (!ret)
-		skb_put(skb, pkt->act_len);
 
 	return ret;
 }
Signature · 7 lines
2.20.1


_______________________________________________
ath10k mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/ath10k
author maintainer
#287 lore raw

PATCH 549/671 RDMA/cma: Fix false error message

In reply to Sasha Levin

From: Håkon Bugge <[email protected]>

[ Upstream commit a6e4d254c19b541a58caced322111084b27a7788 ]

In addr_handler(), assuming status == 0 and the device already has been acquired (id_priv->cma_dev != NULL), we get the following incorrect "error" message:

RDMA CM: ADDR_ERROR: failed to resolve IP. status 0

---
 drivers/infiniband/core/cma.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/core/cma.c +1 -1
index 319bfef00a4a..e16872e0724f 100644
@@ -2889,7 +2889,7 @@ static void addr_handler(int status, struct sockaddr *src_addr,
 		if (status)
 			pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to acquire device. status %d\n",
 					     status);
-	} else {
+	} else if (status) {
 		pr_debug_ratelimited("RDMA CM: ADDR_ERROR: failed to resolve IP. status %d\n", status);
 	}
 
Signature · 1 line
2.20.1
author maintainer
#288 lore raw

PATCH 550/671 RDMA: Fix goto target to release the allocated memory

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit 4a9d46a9fe14401f21df69cea97c62396d5fb053 ]

In bnxt_re_create_srq(), when ib_copy_to_udata() fails allocated memory should be released by goto fail.

Fixes 37cb11acf1f7 ("RDMA/bnxt_re: Add SRQ support for Broadcom adapters")
---
 drivers/infiniband/hw/bnxt_re/ib_verbs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/infiniband/hw/bnxt_re/ib_verbs.c +1 -1
index bea8318e7007..a019ff266f52 100644
@@ -1446,7 +1446,7 @@ struct ib_srq *bnxt_re_create_srq(struct ib_pd *ib_pd,
 			dev_err(rdev_to_dev(rdev), "SRQ copy to udata failed!");
 			bnxt_qplib_destroy_srq(&rdev->qplib_res,
 					       &srq->qplib_srq);
-			goto exit;
+			goto fail;
 		}
 	}
 	if (nq)
Signature · 1 line
2.20.1
author maintainer
#289 lore raw

PATCH 551/671 net/rds: Fix 'ib_evt_handler_call' element in 'rds_ib_stat_names'

In reply to Sasha Levin

From: Gerd Rausch <[email protected]>

[ Upstream commit 05a82481a3024b94db00b8c816bb3d526b5209e0 ]

All entries in 'rds_ib_stat_names' are stringified versions of the corresponding "struct rds_ib_statistics" element without the "s_"-prefix.

Fix entry 'ib_evt_handler_call' to do the same.

Fixes f4f943c958a2 ("RDS: IB: ack more receive completions to improve performance")
---
 net/rds/ib_stats.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/rds/ib_stats.c +1 -1
index 9252ad126335..ac46d8961b61 100644
@@ -42,7 +42,7 @@ DEFINE_PER_CPU_SHARED_ALIGNED(struct rds_ib_statistics, rds_ib_stats);
 static const char *const rds_ib_stat_names[] = {
 	"ib_connect_raced",
 	"ib_listen_closed_stale",
-	"s_ib_evt_handler_call",
+	"ib_evt_handler_call",
 	"ib_tasklet_call",
 	"ib_tx_cq_event",
 	"ib_tx_ring_full",
Signature · 1 line
2.20.1
author maintainer
#290 lore raw

PATCH 552/671 um: Fix off by one error in IRQ enumeration

In reply to Sasha Levin

From: Anton Ivanov <[email protected]>

[ Upstream commit 09ccf0364ca3e94aba4093707ef433ea8014e2a4 ]

Fix an off-by-one in IRQ enumeration

Fixes 49da7e64f33e ("High Performance UML Vector Network Driver")

Reported by: Dana Johnson <[email protected]>

---
 arch/um/include/asm/irq.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/um/include/asm/irq.h +1 -1
index 49ed3e35b35a..ce7a78c3bcf2 100644
@@ -23,7 +23,7 @@
 #define VECTOR_BASE_IRQ		15
 #define VECTOR_IRQ_SPACE	8
 
-#define LAST_IRQ (VECTOR_IRQ_SPACE + VECTOR_BASE_IRQ)
+#define LAST_IRQ (VECTOR_IRQ_SPACE + VECTOR_BASE_IRQ - 1)
 
 #else
 
Signature · 1 line
2.20.1
author maintainer
#291 lore raw

PATCH 553/671 bnxt_en: Increase timeout for HWRM_DBG_COREDUMP_XX commands

In reply to Sasha Levin

From: Vasundhara Volam <[email protected]>

[ Upstream commit 57a8730b1f7a0be7bf8a0a0bb665329074ba764f ]

Firmware coredump messages take much longer than standard messages, so increase the timeout accordingly.

Fixes 6c5657d085ae ("bnxt_en: Add support for ethtool get dump.")
---
 drivers/net/ethernet/broadcom/bnxt/bnxt.h         | 1 +
 drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c | 2 +-
 2 files changed, 2 insertions(+), 1 deletion(-)
drivers/net/ethernet/broadcom/bnxt/bnxt.h +1 -0
index f9e253b705ec..585f5aef0a45 100644
@@ -527,6 +527,7 @@ struct rx_tpa_end_cmp_ext {
 #define DFLT_HWRM_CMD_TIMEOUT		500
 #define HWRM_CMD_TIMEOUT		(bp->hwrm_cmd_timeout)
 #define HWRM_RESET_TIMEOUT		((HWRM_CMD_TIMEOUT) * 4)
+#define HWRM_COREDUMP_TIMEOUT		((HWRM_CMD_TIMEOUT) * 12)
 #define HWRM_RESP_ERR_CODE_MASK		0xffff
 #define HWRM_RESP_LEN_OFFSET		4
 #define HWRM_RESP_LEN_MASK		0xffff0000
drivers/net/ethernet/broadcom/bnxt/bnxt_ethtool.c +1 -1
index cdbb8940a4ae..047024717d65 100644
@@ -2833,7 +2833,7 @@ static int bnxt_hwrm_dbg_coredump_initiate(struct bnxt *bp, u16 component_id,
 	req.component_id = cpu_to_le16(component_id);
 	req.segment_id = cpu_to_le16(segment_id);
 
-	return hwrm_send_message(bp, &req, sizeof(req), HWRM_CMD_TIMEOUT);
+	return hwrm_send_message(bp, &req, sizeof(req), HWRM_COREDUMP_TIMEOUT);
 }
 
 static int bnxt_hwrm_dbg_coredump_retrieve(struct bnxt *bp, u16 component_id,
Signature · 1 line
2.20.1
author maintainer
#292 lore raw

PATCH 554/671 f2fs: fix to avoid accessing uninitialized field of inode page in is_alive()

In reply to Sasha Levin

From: Chao Yu <[email protected]>

[ Upstream commit 98194030554cd9b10568a9b58f5a135c7e7cba85 ]

If inode is newly created, inode page may not synchronize with inode cache, so fields like .i_inline or .i_extra_isize could be wrong, in below call path, we may access such wrong fields, result in failing to migrate valid target block.

Thread A Thread B
- f2fs_create
- f2fs_add_link
- f2fs_add_dentry
- f2fs_init_inode_metadata
- f2fs_add_inline_entry
- f2fs_new_inode_page
- f2fs_put_page
: inode page wasn't updated with inode cache
- gc_data_segment
- is_alive
- f2fs_get_node_page
- datablock_addr
- offset_in_addr : access uninitialized fields

Fixes 7a2af766af15 ("f2fs: enhance on-disk inode structure scalability")
---
 fs/f2fs/dir.c    | 5 +++++
 fs/f2fs/inline.c | 5 +++++
 2 files changed, 10 insertions(+)
fs/f2fs/dir.c +5 -0
index cd611a57d04d..8692cfa89a1c 100644
@@ -572,6 +572,11 @@ int f2fs_add_regular_entry(struct inode *dir, const struct qstr *new_name,
 
 	if (inode) {
 		f2fs_i_pino_write(inode, dir->i_ino);
+
+		/* synchronize inode page's data from inode cache */
+		if (is_inode_flag_set(inode, FI_NEW_INODE))
+			f2fs_update_inode(inode, page);
+
 		f2fs_put_page(page, 1);
 	}
 
fs/f2fs/inline.c +5 -0
index 3fe0dd531390..c1ba29d10789 100644
@@ -578,6 +578,11 @@ int f2fs_add_inline_entry(struct inode *dir, const struct qstr *new_name,
 	/* we don't need to mark_inode_dirty now */
 	if (inode) {
 		f2fs_i_pino_write(inode, dir->i_ino);
+
+		/* synchronize inode page's data from inode cache */
+		if (is_inode_flag_set(inode, FI_NEW_INODE))
+			f2fs_update_inode(inode, page);
+
 		f2fs_put_page(page, 1);
 	}
 
Signature · 8 lines
2.20.1



_______________________________________________
Linux-f2fs-devel mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/linux-f2fs-devel
author maintainer
#293 lore raw

PATCH 555/671 mailbox: qcom-apcs: fix max_register value

In reply to Sasha Levin

From: Jorge Ramirez-Ortiz <[email protected]>

[ Upstream commit 556a0964e28c4441dcdd50fb07596fd042246bd5 ]

The mailbox length is 0x1000 hence the max_register value is 0xFFC.

Fixes c6a8b171ca8e ("mailbox: qcom: Convert APCS IPC driver to use

regmap")

---
 drivers/mailbox/qcom-apcs-ipc-mailbox.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mailbox/qcom-apcs-ipc-mailbox.c +1 -1
index 5255dcb551a7..d8b4f08f613b 100644
@@ -36,7 +36,7 @@ static const struct regmap_config apcs_regmap_config = {
 	.reg_bits = 32,
 	.reg_stride = 4,
 	.val_bits = 32,
-	.max_register = 0x1000,
+	.max_register = 0xFFC,
 	.fast_io = true,
 };
 
Signature · 1 line
2.20.1
author maintainer
#294 lore raw

PATCH 556/671 clk: actions: Fix factor clk struct member access

In reply to Sasha Levin

From: Manivannan Sadhasivam <[email protected]>

[ Upstream commit ed309bfb4812e8b31a3eb877e157b8028a49e50c ]

Since the helper "owl_factor_helper_round_rate" is shared between factor and composite clocks, using the factor clk specific helper function like "hw_to_owl_factor" to access its members will create issues when called from composite clk specific code. Hence, pass the "factor_hw" struct pointer directly instead of fetching it using factor clk specific helpers.

This issue has been observed when a composite clock like "sd0_clk" tried to call "owl_factor_helper_round_rate" resulting in pointer dereferencing error.

While we are at it, let's rename the "clk_val_best" function to "owl_clk_val_best" since this is an owl SoCs specific helper.

---
 drivers/clk/actions/owl-factor.c | 7 +++----
 1 file changed, 3 insertions(+), 4 deletions(-)
drivers/clk/actions/owl-factor.c +3 -4
index 317d4a9e112e..f15e2621fa18 100644
@@ -64,11 +64,10 @@ static unsigned int _get_table_val(const struct clk_factor_table *table,
 	return val;
 }
 
-static int clk_val_best(struct clk_hw *hw, unsigned long rate,
+static int owl_clk_val_best(const struct owl_factor_hw *factor_hw,
+			struct clk_hw *hw, unsigned long rate,
 			unsigned long *best_parent_rate)
 {
-	struct owl_factor *factor = hw_to_owl_factor(hw);
-	struct owl_factor_hw *factor_hw = &factor->factor_hw;
 	const struct clk_factor_table *clkt = factor_hw->table;
 	unsigned long parent_rate, try_parent_rate, best = 0, cur_rate;
 	unsigned long parent_rate_saved = *best_parent_rate;
@@ -126,7 +125,7 @@ long owl_factor_helper_round_rate(struct owl_clk_common *common,
 	const struct clk_factor_table *clkt = factor_hw->table;
 	unsigned int val, mul = 0, div = 1;
 
-	val = clk_val_best(&common->hw, rate, parent_rate);
+	val = owl_clk_val_best(factor_hw, &common->hw, rate, parent_rate);
 	_get_table_div_mul(clkt, val, &mul, &div);
 
 	return *parent_rate * mul / div;
Signature · 1 line
2.20.1
author maintainer
#295 lore raw

PATCH 557/671 powerpc/mm/mce: Keep irqs disabled during lockless page table walk

In reply to Sasha Levin

From: "Aneesh Kumar K.V" <[email protected]>

[ Upstream commit d9101bfa6adc831bda8836c4d774820553c14942 ]

__find_linux_mm_pte() returns a page table entry pointer after walking the page table without holding locks. To make it safe against a THP split and/or collapse, we disable interrupts around the lockless page table walk. However we need to keep interrupts disabled as long as we use the page table entry pointer that is returned.

Fix addr_to_pfn() to do that.

Fixes ba41e1e1ccb9 ("powerpc/mce: Hookup derror (load/store) UE errors")
[mpe: Rearrange code slightly and tweak change log wording]
---
 arch/powerpc/kernel/mce_power.c | 20 ++++++++++++--------
 1 file changed, 12 insertions(+), 8 deletions(-)
arch/powerpc/kernel/mce_power.c +12 -8
index 37a110b8e7e1..ecb375040637 100644
@@ -40,7 +40,7 @@ static unsigned long addr_to_pfn(struct pt_regs *regs, unsigned long addr)
 {
 	pte_t *ptep;
 	unsigned int shift;
-	unsigned long flags;
+	unsigned long pfn, flags;
 	struct mm_struct *mm;
 
 	if (user_mode(regs))
@@ -50,18 +50,22 @@ static unsigned long addr_to_pfn(struct pt_regs *regs, unsigned long addr)
 
 	local_irq_save(flags);
 	ptep = __find_linux_pte(mm->pgd, addr, NULL, &shift);
-	local_irq_restore(flags);
 
-	if (!ptep || pte_special(*ptep))
-		return ULONG_MAX;
+	if (!ptep || pte_special(*ptep)) {
+		pfn = ULONG_MAX;
+		goto out;
+	}
 
-	if (shift > PAGE_SHIFT) {
+	if (shift <= PAGE_SHIFT)
+		pfn = pte_pfn(*ptep);
+	else {
 		unsigned long rpnmask = (1ul << shift) - PAGE_SIZE;
-
-		return pte_pfn(__pte(pte_val(*ptep) | (addr & rpnmask)));
+		pfn = pte_pfn(__pte(pte_val(*ptep) | (addr & rpnmask)));
 	}
 
-	return pte_pfn(*ptep);
+out:
+	local_irq_restore(flags);
+	return pfn;
 }
 
 /* flush SLBs and reload */
Signature · 1 line
2.20.1
author maintainer
#296 lore raw

PATCH 558/671 bpf: fix BTF limits

In reply to Sasha Levin

From: Alexei Starovoitov <[email protected]>

[ Upstream commit a0791f0df7d212c245761538b17a9ea93607b667 ]

vmlinux BTF has more than 64k types.
Its string section is also at the offset larger than 64k.
Adjust both limits to make in-kernel BTF verifier successfully parse in-kernel BTF.

Fixes 69b693f0aefa ("bpf: btf: Introduce BPF Type Format (BTF)")
---
 include/uapi/linux/btf.h | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
include/uapi/linux/btf.h +2 -2
index 972265f32871..1e2662ff0529 100644
@@ -22,9 +22,9 @@ struct btf_header {
 };
 
 /* Max # of type identifier */
-#define BTF_MAX_TYPE	0x0000ffff
+#define BTF_MAX_TYPE	0x000fffff
 /* Max offset into the string section */
-#define BTF_MAX_NAME_OFFSET	0x0000ffff
+#define BTF_MAX_NAME_OFFSET	0x00ffffff
 /* Max # of struct/union/enum members or func args */
 #define BTF_MAX_VLEN	0xffff
 
Signature · 1 line
2.20.1
author maintainer
#297 lore raw

PATCH 559/671 crypto: hisilicon - Matching the dma address for dma_pool_free()

In reply to Sasha Levin

From: Yunfeng Ye <[email protected]>

[ Upstream commit e00371af1d4ce73d527d8ee69fda2febaf5a42c2 ]

When dma_pool_zalloc() fail in sec_alloc_and_fill_hw_sgl(), dma_pool_free() is invoked, but the parameters that sgl_current and sgl_current->next_sgl is not match.

Using sec_free_hw_sgl() instead of the original free routine.

Fixes 915e4e8413da ("crypto: hisilicon - SEC security accelerator driver")
---
 drivers/crypto/hisilicon/sec/sec_algs.c | 44 +++++++++++--------------
 1 file changed, 19 insertions(+), 25 deletions(-)
drivers/crypto/hisilicon/sec/sec_algs.c +19 -25
index db2983c51f1e..bf9658800bda 100644
@@ -153,6 +153,24 @@ static void sec_alg_skcipher_init_context(struct crypto_skcipher *atfm,
 				       ctx->cipher_alg);
 }
 
+static void sec_free_hw_sgl(struct sec_hw_sgl *hw_sgl,
+			    dma_addr_t psec_sgl, struct sec_dev_info *info)
+{
+	struct sec_hw_sgl *sgl_current, *sgl_next;
+	dma_addr_t sgl_next_dma;
+
+	sgl_current = hw_sgl;
+	while (sgl_current) {
+		sgl_next = sgl_current->next;
+		sgl_next_dma = sgl_current->next_sgl;
+
+		dma_pool_free(info->hw_sgl_pool, sgl_current, psec_sgl);
+
+		sgl_current = sgl_next;
+		psec_sgl = sgl_next_dma;
+	}
+}
+
 static int sec_alloc_and_fill_hw_sgl(struct sec_hw_sgl **sec_sgl,
 				     dma_addr_t *psec_sgl,
 				     struct scatterlist *sgl,
@@ -199,36 +217,12 @@ static int sec_alloc_and_fill_hw_sgl(struct sec_hw_sgl **sec_sgl,
 	return 0;
 
 err_free_hw_sgls:
-	sgl_current = *sec_sgl;
-	while (sgl_current) {
-		sgl_next = sgl_current->next;
-		dma_pool_free(info->hw_sgl_pool, sgl_current,
-			      sgl_current->next_sgl);
-		sgl_current = sgl_next;
-	}
+	sec_free_hw_sgl(*sec_sgl, *psec_sgl, info);
 	*psec_sgl = 0;
 
 	return ret;
 }
 
-static void sec_free_hw_sgl(struct sec_hw_sgl *hw_sgl,
-			    dma_addr_t psec_sgl, struct sec_dev_info *info)
-{
-	struct sec_hw_sgl *sgl_current, *sgl_next;
-	dma_addr_t sgl_next_dma;
-
-	sgl_current = hw_sgl;
-	while (sgl_current) {
-		sgl_next = sgl_current->next;
-		sgl_next_dma = sgl_current->next_sgl;
-
-		dma_pool_free(info->hw_sgl_pool, sgl_current, psec_sgl);
-
-		sgl_current = sgl_next;
-		psec_sgl = sgl_next_dma;
-	}
-}
-
 static int sec_alg_skcipher_setkey(struct crypto_skcipher *tfm,
 				   const u8 *key, unsigned int keylen,
 				   enum sec_cipher_alg alg)
Signature · 1 line
2.20.1
author maintainer
#298 lore raw

PATCH 560/671 iommu/amd: Wait for completion of IOTLB flush in attach_device

In reply to Sasha Levin

From: Filippo Sironi <[email protected]>

[ Upstream commit 0b15e02f0cc4fb34a9160de7ba6db3a4013dc1b7 ]

To make sure the domain tlb flush completes before the function returns, explicitly wait for its completion.

Fixes 42a49f965a8d ("amd-iommu: flush domain tlb when attaching a new device")
[joro: Added commit message and fixes tag]
---
 drivers/iommu/amd_iommu.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/iommu/amd_iommu.c +2 -0
index 9991386fb700..bea19aa33758 100644
@@ -2153,6 +2153,8 @@ static int attach_device(struct device *dev,
 	 */
 	domain_flush_tlb_pde(domain);
 
+	domain_flush_complete(domain);
+
 	return ret;
 }
 
Signature · 6 lines
2.20.1

_______________________________________________
iommu mailing list
[email protected]
https://lists.linuxfoundation.org/mailman/listinfo/iommu
author maintainer
#299 lore raw

PATCH 561/671 net: aquantia: Fix aq_vec_isr_legacy() return value

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 31aefe14bc9f56566041303d733fda511d3a1c3e ]

The irqreturn_t type is an enum or an unsigned int in GCC. That creates to problems because it can't detect if the self->aq_hw_ops->hw_irq_read() call fails and at the end the function always returns IRQ_HANDLED.

drivers/net/ethernet/aquantia/atlantic/aq_vec.c:316 aq_vec_isr_legacy() warn: unsigned 'err' is never less than zero. drivers/net/ethernet/aquantia/atlantic/aq_vec.c:329 aq_vec_isr_legacy() warn: always true condition '(err >= 0) => (0-u32max >= 0)'

Fixes 970a2e9864b0 ("net: ethernet: aquantia: Vector operations")
---
 drivers/net/ethernet/aquantia/atlantic/aq_vec.c | 15 ++++++---------
 1 file changed, 6 insertions(+), 9 deletions(-)
drivers/net/ethernet/aquantia/atlantic/aq_vec.c +6 -9
index 82582fa54d5d..72aa2a4c4d66 100644
@@ -309,15 +309,13 @@ irqreturn_t aq_vec_isr_legacy(int irq, void *private)
 {
 	struct aq_vec_s *self = private;
 	u64 irq_mask = 0U;
-	irqreturn_t err = 0;
+	int err;
 
-	if (!self) {
-		err = -EINVAL;
-		goto err_exit;
-	}
+	if (!self)
+		return IRQ_NONE;
 	err = self->aq_hw_ops->hw_irq_read(self->aq_hw, &irq_mask);
 	if (err < 0)
-		goto err_exit;
+		return IRQ_NONE;
 
 	if (irq_mask) {
 		self->aq_hw_ops->hw_irq_disable(self->aq_hw,
@@ -325,11 +323,10 @@ irqreturn_t aq_vec_isr_legacy(int irq, void *private)
 		napi_schedule(&self->napi);
 	} else {
 		self->aq_hw_ops->hw_irq_enable(self->aq_hw, 1U);
-		err = IRQ_NONE;
+		return IRQ_NONE;
 	}
 
-err_exit:
-	return err >= 0 ? IRQ_HANDLED : IRQ_NONE;
+	return IRQ_HANDLED;
 }
 
 cpumask_t *aq_vec_get_affinity_mask(struct aq_vec_s *self)
Signature · 1 line
2.20.1
author maintainer
#300 lore raw

PATCH 562/671 cxgb4: Signedness bug in init_one()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 286183147666fb76c057836c57d86e9e6f508bca ]

The "chip" variable is an enum, and it's treated as unsigned int by GCC in this context so the error handling isn't triggered.

Fixes e8d452923ae6 ("cxgb4: clean up init_one")
---
 drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/chelsio/cxgb4/cxgb4_main.c +1 -1
index bb04c695ab9f..c81d6c330548 100644
@@ -5452,7 +5452,7 @@ static int init_one(struct pci_dev *pdev, const struct pci_device_id *ent)
 	whoami = t4_read_reg(adapter, PL_WHOAMI_A);
 	pci_read_config_word(pdev, PCI_DEVICE_ID, &device_id);
 	chip = t4_get_chip_type(adapter, CHELSIO_PCI_ID_VER(device_id));
-	if (chip < 0) {
+	if ((int)chip < 0) {
 		dev_err(&pdev->dev, "Device %d is not supported\n", device_id);
 		err = chip;
 		goto out_free_adapter;
Signature · 1 line
2.20.1
author maintainer
#301 lore raw

PATCH 563/671 net: hisilicon: Fix signedness bug in hix5hd2_dev_probe()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 002dfe8085255b7bf1e0758c3d195c5412d35be9 ]

The "priv->phy_mode" variable is an enum and in this context GCC will treat it as unsigned to the error handling will never trigger.

Fixes 57c5bc9ad7d7 ("net: hisilicon: add hix5hd2 mac driver")
---
 drivers/net/ethernet/hisilicon/hix5hd2_gmac.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/hisilicon/hix5hd2_gmac.c +1 -1
index 471805ea363b..b63871ef8a40 100644
@@ -1201,7 +1201,7 @@ static int hix5hd2_dev_probe(struct platform_device *pdev)
 		goto err_free_mdio;
 
 	priv->phy_mode = of_get_phy_mode(node);
-	if (priv->phy_mode < 0) {
+	if ((int)priv->phy_mode < 0) {
 		netdev_err(ndev, "not find phy-mode\n");
 		ret = -EINVAL;
 		goto err_mdiobus;
Signature · 1 line
2.20.1
author maintainer
#302 lore raw

PATCH 564/671 net: broadcom/bcmsysport: Fix signedness in bcm_sysport_probe()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 25a584955f020d6ec499c513923fb220f3112d2b ]

The "priv->phy_interface" variable is an enum and in this context GCC will treat it as unsigned so the error handling will never be triggered.

Fixes 80105befdb4b ("net: systemport: add Broadcom SYSTEMPORT Ethernet MAC driver")
---
 drivers/net/ethernet/broadcom/bcmsysport.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/broadcom/bcmsysport.c +1 -1
index 0bdbc72605e1..49aa3b5ea57c 100644
@@ -2470,7 +2470,7 @@ static int bcm_sysport_probe(struct platform_device *pdev)
 
 	priv->phy_interface = of_get_phy_mode(dn);
 	/* Default to GMII interface mode */
-	if (priv->phy_interface < 0)
+	if ((int)priv->phy_interface < 0)
 		priv->phy_interface = PHY_INTERFACE_MODE_GMII;
 
 	/* In the case of a fixed PHY, the DT node associated
Signature · 1 line
2.20.1
author maintainer
#303 lore raw

PATCH 565/671 net: netsec: Fix signedness bug in netsec_probe()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit bd55f8ddbc437c225391ca8f487e7ec10243c4cc ]

The "priv->phy_interface" variable is an enum and in this context GCC will treat it as an unsigned int so the error handling is never triggered.

Fixes 533dd11a12f6 ("net: socionext: Add Synquacer NetSec driver")
---
 drivers/net/ethernet/socionext/netsec.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/socionext/netsec.c +1 -1
index d9d0d03e4ce7..027367b9cc48 100644
@@ -1604,7 +1604,7 @@ static int netsec_probe(struct platform_device *pdev)
 			   NETIF_MSG_LINK | NETIF_MSG_PROBE;
 
 	priv->phy_interface = device_get_phy_mode(&pdev->dev);
-	if (priv->phy_interface < 0) {
+	if ((int)priv->phy_interface < 0) {
 		dev_err(&pdev->dev, "missing required property 'phy-mode'\n");
 		ret = -ENODEV;
 		goto free_ndev;
Signature · 1 line
2.20.1
author maintainer
#304 lore raw

PATCH 566/671 net: socionext: Fix a signedness bug in ave_probe()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 7f9e88e6ef8c971f2c638b5ff7044c59b5d0f58d ]

The "phy_mode" variable is an enum and in this context GCC treats it as an unsigned int so the error handling is never triggered.

Fixes 4c270b55a5af ("net: ethernet: socionext: add AVE ethernet driver")
---
 drivers/net/ethernet/socionext/sni_ave.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/socionext/sni_ave.c +1 -1
index 09d25b87cf7c..c309accc6797 100644
@@ -1575,7 +1575,7 @@ static int ave_probe(struct platform_device *pdev)
 
 	np = dev->of_node;
 	phy_mode = of_get_phy_mode(np);
-	if (phy_mode < 0) {
+	if ((int)phy_mode < 0) {
 		dev_err(dev, "phy-mode not found\n");
 		return -EINVAL;
 	}
Signature · 1 line
2.20.1
author maintainer
#305 lore raw

PATCH 567/671 net: stmmac: dwmac-meson8b: Fix signedness bug in probe

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit f10210517a2f37feea2edf85eb34c98977265c16 ]

The "dwmac->phy_mode" is an enum and in this context GCC treats it as an unsigned int so the error handling is never triggered.

Fixes 566e82516253 ("net: stmmac: add a glue driver for the Amlogic Meson 8b / GXBB DWMAC")
---
 drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/dwmac-meson8b.c +1 -1
index 94b46258e8ff..0a17535f13ae 100644
@@ -355,7 +355,7 @@ static int meson8b_dwmac_probe(struct platform_device *pdev)
 
 	dwmac->dev = &pdev->dev;
 	dwmac->phy_mode = of_get_phy_mode(pdev->dev.of_node);
-	if (dwmac->phy_mode < 0) {
+	if ((int)dwmac->phy_mode < 0) {
 		dev_err(&pdev->dev, "missing phy-mode property\n");
 		ret = -EINVAL;
 		goto err_remove_config_dt;
Signature · 1 line
2.20.1
author maintainer
#306 lore raw

PATCH 568/671 net: axienet: fix a signedness bug in probe

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 73e211e11be86715d66bd3c9d38b3c34b05fca9a ]

The "lp->phy_mode" is an enum but in this context GCC treats it as an unsigned int so the error handling is never triggered.

Fixes ee06b1728b95 ("net: axienet: add support for standard phy-mode binding")
---
 drivers/net/ethernet/xilinx/xilinx_axienet_main.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/xilinx/xilinx_axienet_main.c +1 -1
index 28764268a44f..b093f14eeec3 100644
@@ -1573,7 +1573,7 @@ static int axienet_probe(struct platform_device *pdev)
 		}
 	} else {
 		lp->phy_mode = of_get_phy_mode(pdev->dev.of_node);
-		if (lp->phy_mode < 0) {
+		if ((int)lp->phy_mode < 0) {
 			ret = -EINVAL;
 			goto free_netdev;
 		}
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#307 lore raw

PATCH 569/671 of: mdio: Fix a signedness bug in of_phy_get_and_connect()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit d7eb651212fdbafa82d485d8e76095ac3b14c193 ]

The "iface" variable is an enum and in this context GCC treats it as an unsigned int so the error handling is never triggered.

Fixes b78624125304 ("of_mdio: Abstract a general interface for phy connect")
---
 drivers/of/of_mdio.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/of/of_mdio.c +1 -1
index 5ad1342f5682..7d2bc22680d9 100644
@@ -370,7 +370,7 @@ struct phy_device *of_phy_get_and_connect(struct net_device *dev,
 	int ret;
 
 	iface = of_get_phy_mode(np);
-	if (iface < 0)
+	if ((int)iface < 0)
 		return NULL;
 	if (of_phy_is_fixed_link(np)) {
 		ret = of_phy_register_fixed_link(np);
Signature · 1 line
2.20.1
author maintainer
#308 lore raw

PATCH 570/671 net: nixge: Fix a signedness bug in nixge_probe()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 1a4b62a0b8a3b81eca24366f63e214a7144b9f02 ]

The "priv->phy_mode" is an enum and in this context GCC will treat it as an unsigned int so it can never be less than zero.

Fixes 492caffa8a1a ("net: ethernet: nixge: Add support for National Instruments XGE netdev")
---
 drivers/net/ethernet/ni/nixge.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/ni/nixge.c +1 -1
index 76efed058f33..a791d7932b0e 100644
@@ -1233,7 +1233,7 @@ static int nixge_probe(struct platform_device *pdev)
 	}
 
 	priv->phy_mode = of_get_phy_mode(pdev->dev.of_node);
-	if (priv->phy_mode < 0) {
+	if ((int)priv->phy_mode < 0) {
 		netdev_err(ndev, "not find \"phy-mode\" property\n");
 		err = -EINVAL;
 		goto unregister_mdio;
Signature · 1 line
2.20.1
author maintainer
#309 lore raw

PATCH 571/671 net: ethernet: stmmac: Fix signedness bug in ipq806x_gmac_of_parse()

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 231042181dc9d6122c6faba64e99ccb25f13cc6c ]

The "gmac->phy_mode" variable is an enum and in this context GCC will treat it as an unsigned int so the error handling will never be triggered.

Fixes b1c17215d718 ("stmmac: add ipq806x glue layer")
---
 drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/dwmac-ipq806x.c +1 -1
index 2c6d7c69c8f7..0d21082ceb93 100644
@@ -191,7 +191,7 @@ static int ipq806x_gmac_of_parse(struct ipq806x_gmac *gmac)
 	struct device *dev = &gmac->pdev->dev;
 
 	gmac->phy_mode = of_get_phy_mode(dev->of_node);
-	if (gmac->phy_mode < 0) {
+	if ((int)gmac->phy_mode < 0) {
 		dev_err(dev, "missing phy mode property\n");
 		return -EINVAL;
 	}
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#310 lore raw

PATCH 572/671 net: sched: cbs: Avoid division by zero when calculating the port rate

In reply to Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 83c8c3cf45163f0c823db37be6ab04dfcf8ac751 ]

As explained in the "net: sched: taprio: Avoid division by zero on invalid link speed" commit, it is legal for the ethtool API to return zero as a link speed. So guard against it to ensure we don't perform a division by zero in kernel.

Fixes e0a7683d30e9 ("net/sched: cbs: fix port_rate miscalculation")
---
 net/sched/sch_cbs.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
net/sched/sch_cbs.c +1 -1
index 289f66b9238d..940e72d6db18 100644
@@ -310,7 +310,7 @@ static void cbs_set_port_rate(struct net_device *dev, struct cbs_sched_data *q)
 	if (err < 0)
 		goto skip;
 
-	if (ecmd.base.speed != SPEED_UNKNOWN)
+	if (ecmd.base.speed && ecmd.base.speed != SPEED_UNKNOWN)
 		speed = ecmd.base.speed;
 
 skip:
Signature · 1 line
2.20.1
author maintainer
#311 lore raw

PATCH 573/671 ipv6: Handle race in addrconf_dad_work

In reply to Sasha Levin

From: David Ahern <[email protected]>

[ Upstream commit a3ce2a21bb8969ae27917281244fa91bf5f286d7 ]

Rajendra reported a kernel panic when a link was taken down:

[ 6870.263084] BUG: unable to handle kernel NULL pointer dereference at 00000000000000a8
[ 6870.271856] IP: [<ffffffff8efc5764>] __ipv6_ifa_notify+0x154/0x290

<snip>

[ 6870.570501] Call Trace:
[ 6870.573238] [<ffffffff8efc58c6>] ? ipv6_ifa_notify+0x26/0x40
[ 6870.579665] [<ffffffff8efc98ec>] ? addrconf_dad_completed+0x4c/0x2c0
[ 6870.586869] [<ffffffff8efe70c6>] ? ipv6_dev_mc_inc+0x196/0x260
[ 6870.593491] [<ffffffff8efc9c6a>] ? addrconf_dad_work+0x10a/0x430
[ 6870.600305] [<ffffffff8f01ade4>] ? __switch_to_asm+0x34/0x70
[ 6870.606732] [<ffffffff8ea93a7a>] ? process_one_work+0x18a/0x430
[ 6870.613449] [<ffffffff8ea93d6d>] ? worker_thread+0x4d/0x490
[ 6870.619778] [<ffffffff8ea93d20>] ? process_one_work+0x430/0x430
[ 6870.626495] [<ffffffff8ea99dd9>] ? kthread+0xd9/0xf0
[ 6870.632145] [<ffffffff8f01ade4>] ? __switch_to_asm+0x34/0x70
[ 6870.638573] [<ffffffff8ea99d00>] ? kthread_park+0x60/0x60
[ 6870.644707] [<ffffffff8f01ae77>] ? ret_from_fork+0x57/0x70
[ 6870.650936] Code: 31 c0 31 d2 41 b9 20 00 08 02 b9 09 00 00 0

addrconf_dad_work is kicked to be scheduled when a device is brought up. There is a race between addrcond_dad_work getting scheduled and taking the rtnl lock and a process taking the link down (under rtnl). The latter removes the host route from the inet6_addr as part of addrconf_ifdown which is run for NETDEV_DOWN. The former attempts to use the host route in ipv6_ifa_notify. If the down event removes the host route due to the race to the rtnl, then the BUG listed above occurs.

This scenario does not occur when the ipv6 address is not kept (net.ipv6.conf.all.keep_addr_on_down = 0) as addrconf_ifdown sets the state of the ifp to DEAD. Handle when the addresses are kept by checking IF_READY which is reset by addrconf_ifdown.

The 'dead' flag for an inet6_addr is set only under rtnl, in addrconf_ifdown and it means the device is getting removed (or IPv6 is disabled). The interesting cases for changing the idev flag are addrconf_notify (NETDEV_UP and NETDEV_CHANGE) and addrconf_ifdown (reset the flag). The former does not have the idev lock - only rtnl; the latter has both. Based on that the existing dead + IF_READY check can be moved to right after the rtnl_lock in addrconf_dad_work.

Fixes f1705ec197e7 ("net: ipv6: Make address flushing on ifdown optional")
---
 net/ipv6/addrconf.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
net/ipv6/addrconf.c +6 -5
index ef309a26aba0..4cb94b5a9d39 100644
@@ -3978,6 +3978,12 @@ static void addrconf_dad_work(struct work_struct *w)
 
 	rtnl_lock();
 
+	/* check if device was taken down before this delayed work
+	 * function could be canceled
+	 */
+	if (idev->dead || !(idev->if_flags & IF_READY))
+		goto out;
+
 	spin_lock_bh(&ifp->lock);
 	if (ifp->state == INET6_IFADDR_STATE_PREDAD) {
 		action = DAD_BEGIN;
@@ -4023,11 +4029,6 @@ static void addrconf_dad_work(struct work_struct *w)
 		goto out;
 
 	write_lock_bh(&idev->lock);
-	if (idev->dead || !(idev->if_flags & IF_READY)) {
-		write_unlock_bh(&idev->lock);
-		goto out;
-	}
-
 	spin_lock(&ifp->lock);
 	if (ifp->state == INET6_IFADDR_STATE_DEAD) {
 		spin_unlock(&ifp->lock);
Signature · 1 line
2.20.1
#312 lore raw

Re: [PATCH AUTOSEL 4.19 573/671] ipv6: Handle race in addrconf_dad_work

Quoted 4 lines · Sasha Levin

From: David Ahern <[email protected]>

[ Upstream commit a3ce2a21bb8969ae27917281244fa91bf5f286d7 ]

That commit was reverted by 8ae72cbf62d2c1879456c0c5872f958e18f53711 and then replaced by 2d819d250a1393a3e725715425ab70a0e0772a71

#313 lore raw

Re: [PATCH AUTOSEL 4.19 573/671] ipv6: Handle race in addrconf_dad_work

Quoted 9 lines · David Ahern
Quoted 4 lines · Sasha Levin

From: David Ahern <[email protected]>

[ Upstream commit a3ce2a21bb8969ae27917281244fa91bf5f286d7 ]

That commit was reverted by 8ae72cbf62d2c1879456c0c5872f958e18f53711 and then replaced by 2d819d250a1393a3e725715425ab70a0e0772a71

BTW, the AUTOSEL algorithm should be updated to look for reverts and even ones that have already been nack'ed from a backport perspective.

I felt a bit of deja vu with my response and sure enough this patch was selected back in October and I responded then that it should not be backported.

author maintainer
#314 lore raw

Re: [PATCH AUTOSEL 4.19 573/671] ipv6: Handle race in addrconf_dad_work

Quoted 11 lines · David Ahern
Quoted 9 lines · David Ahern
Quoted 4 lines · Sasha Levin

From: David Ahern <[email protected]>

[ Upstream commit a3ce2a21bb8969ae27917281244fa91bf5f286d7 ]

That commit was reverted by 8ae72cbf62d2c1879456c0c5872f958e18f53711 and then replaced by 2d819d250a1393a3e725715425ab70a0e0772a71

Quoted 6 lines

BTW, the AUTOSEL algorithm should be updated to look for reverts and even ones that have already been nack'ed from a backport perspective.

I felt a bit of deja vu with my response and sure enough this patch was selected back in October and I responded then that it should not be backported.

Sorry about this David. This series is a result of an experimental work I did rather than the regular AUTOSEL workflow, so it ended up accidentally bubbling a few commits that were previously rejected.

Signature · 2 lines
Thanks,
Sasha
author maintainer
#315 lore raw

PATCH 574/671 nvme: retain split access workaround for capability reads

In reply to Sasha Levin

From: Ard Biesheuvel <[email protected]>

[ Upstream commit 3a8ecc935efabdad106b5e06d07b150c394b4465 ]

Commit 7fd8930f26be4

"nvme: add a common helper to read Identify Controller data"

has re-introduced an issue that we have attempted to work around in the past, in commit a310acd7a7ea ("NVMe: use split lo_hi_{read,write}q").

The problem is that some PCIe NVMe controllers do not implement 64-bit outbound accesses correctly, which is why the commit above switched to using lo_hi_[read|write]q for all 64-bit BAR accesses occuring in the code.

In the mean time, the NVMe subsystem has been refactored, and now calls into the PCIe support layer for NVMe via a .reg_read64() method, which fails to use lo_hi_readq(), and thus reintroduces the problem that the workaround above aimed to address.

Given that, at the moment, .reg_read64() is only used to read the capability register [which is known to tolerate split reads], let's switch .reg_read64() to lo_hi_readq() as well.

This fixes a boot issue on some ARM boxes with NVMe behind a Synopsys DesignWare PCIe host controller.

Fixes 7fd8930f26be4 ("nvme: add a common helper to read Identify Controller data")
---
 drivers/nvme/host/pci.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/nvme/host/pci.c +1 -1
index 124f41157173..3c68a5b35ec1 100644
@@ -2396,7 +2396,7 @@ static int nvme_pci_reg_write32(struct nvme_ctrl *ctrl, u32 off, u32 val)
 
 static int nvme_pci_reg_read64(struct nvme_ctrl *ctrl, u32 off, u64 *val)
 {
-	*val = readq(to_nvme_dev(ctrl)->bar + off);
+	*val = lo_hi_readq(to_nvme_dev(ctrl)->bar + off);
 	return 0;
 }
 
Signature · 7 lines
2.20.1


_______________________________________________
linux-nvme mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-nvme
author maintainer
#316 lore raw

PATCH 575/671 net: stmmac: gmac4+: Not all Unicast addresses may be available

In reply to Sasha Levin

From: Jose Abreu <[email protected]>

[ Upstream commit 25683bab09a70542b9f8e3e28f79b3369e56701f ]

Some setups may not have all Unicast addresses filters available. Check the number of available filters before trying to setup it.

Fixes 477286b53f55 ("stmmac: add GMAC4 core support")
---
 drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/dwmac4_core.c +1 -1
index 48cf5e2b2441..bc8871e7351f 100644
@@ -443,7 +443,7 @@ static void dwmac4_set_filter(struct mac_device_info *hw,
 	}
 
 	/* Handle multiple unicast addresses */
-	if (netdev_uc_count(dev) > GMAC_MAX_PERFECT_ADDRESSES) {
+	if (netdev_uc_count(dev) > hw->unicast_filter_entries) {
 		/* Switch to promiscuous mode if more than 128 addrs
 		 * are required
 		 */
Signature · 1 line
2.20.1
author maintainer
#317 lore raw

PATCH 576/671 rxrpc: Fix trace-after-put looking at the put connection record

In reply to Sasha Levin

From: David Howells <[email protected]>

[ Upstream commit 4c1295dccc0afe0905b6ca4c62ade7f2406f2cfb ]

rxrpc_put_*conn() calls trace_rxrpc_conn() after they have done the decrement of the refcount - which looks at the debug_id in the connection record. But unless the refcount was reduced to zero, we no longer have the right to look in the record and, indeed, it may be deleted by some other thread.

Fix this by getting the debug_id out before decrementing the refcount and then passing that into the tracepoint.

Fixes 363deeab6d0f ("rxrpc: Add connection tracepoint and client conn state tracepoint")
---
 include/trace/events/rxrpc.h |  6 +++---
 net/rxrpc/call_accept.c      |  2 +-
 net/rxrpc/conn_client.c      |  6 ++++--
 net/rxrpc/conn_object.c      | 13 +++++++------
 net/rxrpc/conn_service.c     |  2 +-
 5 files changed, 16 insertions(+), 13 deletions(-)
include/trace/events/rxrpc.h +3 -3
index a08916eb7615..0924119bcfa4 100644
@@ -554,10 +554,10 @@ TRACE_EVENT(rxrpc_peer,
 	    );
 
 TRACE_EVENT(rxrpc_conn,
-	    TP_PROTO(struct rxrpc_connection *conn, enum rxrpc_conn_trace op,
+	    TP_PROTO(unsigned int conn_debug_id, enum rxrpc_conn_trace op,
 		     int usage, const void *where),
 
-	    TP_ARGS(conn, op, usage, where),
+	    TP_ARGS(conn_debug_id, op, usage, where),
 
 	    TP_STRUCT__entry(
 		    __field(unsigned int,	conn		)
@@ -567,7 +567,7 @@ TRACE_EVENT(rxrpc_conn,
 			     ),
 
 	    TP_fast_assign(
-		    __entry->conn = conn->debug_id;
+		    __entry->conn = conn_debug_id;
 		    __entry->op = op;
 		    __entry->usage = usage;
 		    __entry->where = where;
net/rxrpc/call_accept.c +1 -1
index 8079aacaecac..c5566bc4aaca 100644
@@ -88,7 +88,7 @@ static int rxrpc_service_prealloc_one(struct rxrpc_sock *rx,
 		smp_store_release(&b->conn_backlog_head,
 				  (head + 1) & (size - 1));
 
-		trace_rxrpc_conn(conn, rxrpc_conn_new_service,
+		trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_service,
 				 atomic_read(&conn->usage), here);
 	}
 
net/rxrpc/conn_client.c +4 -2
index 3dbb126e6060..38d548532024 100644
@@ -217,7 +217,8 @@ rxrpc_alloc_client_connection(struct rxrpc_conn_parameters *cp, gfp_t gfp)
 	rxrpc_get_local(conn->params.local);
 	key_get(conn->params.key);
 
-	trace_rxrpc_conn(conn, rxrpc_conn_new_client, atomic_read(&conn->usage),
+	trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_client,
+			 atomic_read(&conn->usage),
 			 __builtin_return_address(0));
 	trace_rxrpc_client(conn, -1, rxrpc_client_alloc);
 	_leave(" = %p", conn);
@@ -989,11 +990,12 @@ rxrpc_put_one_client_conn(struct rxrpc_connection *conn)
 void rxrpc_put_client_conn(struct rxrpc_connection *conn)
 {
 	const void *here = __builtin_return_address(0);
+	unsigned int debug_id = conn->debug_id;
 	int n;
 
 	do {
 		n = atomic_dec_return(&conn->usage);
-		trace_rxrpc_conn(conn, rxrpc_conn_put_client, n, here);
+		trace_rxrpc_conn(debug_id, rxrpc_conn_put_client, n, here);
 		if (n > 0)
 			return;
 		ASSERTCMP(n, >=, 0);
net/rxrpc/conn_object.c +7 -6
index 004a6eb529bc..f338efd2880a 100644
@@ -272,7 +272,7 @@ bool rxrpc_queue_conn(struct rxrpc_connection *conn)
 	if (n == 0)
 		return false;
 	if (rxrpc_queue_work(&conn->processor))
-		trace_rxrpc_conn(conn, rxrpc_conn_queued, n + 1, here);
+		trace_rxrpc_conn(conn->debug_id, rxrpc_conn_queued, n + 1, here);
 	else
 		rxrpc_put_connection(conn);
 	return true;
@@ -287,7 +287,7 @@ void rxrpc_see_connection(struct rxrpc_connection *conn)
 	if (conn) {
 		int n = atomic_read(&conn->usage);
 
-		trace_rxrpc_conn(conn, rxrpc_conn_seen, n, here);
+		trace_rxrpc_conn(conn->debug_id, rxrpc_conn_seen, n, here);
 	}
 }
 
@@ -299,7 +299,7 @@ void rxrpc_get_connection(struct rxrpc_connection *conn)
 	const void *here = __builtin_return_address(0);
 	int n = atomic_inc_return(&conn->usage);
 
-	trace_rxrpc_conn(conn, rxrpc_conn_got, n, here);
+	trace_rxrpc_conn(conn->debug_id, rxrpc_conn_got, n, here);
 }
 
 /*
@@ -313,7 +313,7 @@ rxrpc_get_connection_maybe(struct rxrpc_connection *conn)
 	if (conn) {
 		int n = atomic_fetch_add_unless(&conn->usage, 1, 0);
 		if (n > 0)
-			trace_rxrpc_conn(conn, rxrpc_conn_got, n + 1, here);
+			trace_rxrpc_conn(conn->debug_id, rxrpc_conn_got, n + 1, here);
 		else
 			conn = NULL;
 	}
@@ -336,10 +336,11 @@ static void rxrpc_set_service_reap_timer(struct rxrpc_net *rxnet,
 void rxrpc_put_service_conn(struct rxrpc_connection *conn)
 {
 	const void *here = __builtin_return_address(0);
+	unsigned int debug_id = conn->debug_id;
 	int n;
 
 	n = atomic_dec_return(&conn->usage);
-	trace_rxrpc_conn(conn, rxrpc_conn_put_service, n, here);
+	trace_rxrpc_conn(debug_id, rxrpc_conn_put_service, n, here);
 	ASSERTCMP(n, >=, 0);
 	if (n == 1)
 		rxrpc_set_service_reap_timer(conn->params.local->rxnet,
@@ -423,7 +424,7 @@ void rxrpc_service_connection_reaper(struct work_struct *work)
 		 */
 		if (atomic_cmpxchg(&conn->usage, 1, 0) != 1)
 			continue;
-		trace_rxrpc_conn(conn, rxrpc_conn_reap_service, 0, NULL);
+		trace_rxrpc_conn(conn->debug_id, rxrpc_conn_reap_service, 0, NULL);
 
 		if (rxrpc_conn_is_client(conn))
 			BUG();
net/rxrpc/conn_service.c +1 -1
index 80773a50c755..6da7c4bf15e8 100644
@@ -138,7 +138,7 @@ struct rxrpc_connection *rxrpc_prealloc_service_connection(struct rxrpc_net *rxn
 		list_add_tail(&conn->proc_link, &rxnet->conn_proc_list);
 		write_unlock(&rxnet->conn_lock);
 
-		trace_rxrpc_conn(conn, rxrpc_conn_new_service,
+		trace_rxrpc_conn(conn->debug_id, rxrpc_conn_new_service,
 				 atomic_read(&conn->usage),
 				 __builtin_return_address(0));
 	}
Signature · 1 line
2.20.1
author maintainer
#318 lore raw

PATCH 577/671 rxrpc: Fix trace-after-put looking at the put call record

In reply to Sasha Levin

From: David Howells <[email protected]>

[ Upstream commit 48c9e0ec7cbbb7370448f859ccc8e3b7eb69e755 ]

rxrpc_put_call() calls trace_rxrpc_call() after it has done the decrement of the refcount - which looks at the debug_id in the call record. But unless the refcount was reduced to zero, we no longer have the right to look in the record and, indeed, it may be deleted by some other thread.

Fix this by getting the debug_id out before decrementing the refcount and then passing that into the tracepoint.

Fixes e34d4234b0b7 ("rxrpc: Trace rxrpc_call usage")
---
 include/trace/events/rxrpc.h |  6 +++---
 net/rxrpc/call_accept.c      |  2 +-
 net/rxrpc/call_object.c      | 28 +++++++++++++++++-----------
 3 files changed, 21 insertions(+), 15 deletions(-)
include/trace/events/rxrpc.h +3 -3
index 0924119bcfa4..d496794a8f59 100644
@@ -614,10 +614,10 @@ TRACE_EVENT(rxrpc_client,
 	    );
 
 TRACE_EVENT(rxrpc_call,
-	    TP_PROTO(struct rxrpc_call *call, enum rxrpc_call_trace op,
+	    TP_PROTO(unsigned int call_debug_id, enum rxrpc_call_trace op,
 		     int usage, const void *where, const void *aux),
 
-	    TP_ARGS(call, op, usage, where, aux),
+	    TP_ARGS(call_debug_id, op, usage, where, aux),
 
 	    TP_STRUCT__entry(
 		    __field(unsigned int,		call		)
@@ -628,7 +628,7 @@ TRACE_EVENT(rxrpc_call,
 			     ),
 
 	    TP_fast_assign(
-		    __entry->call = call->debug_id;
+		    __entry->call = call_debug_id;
 		    __entry->op = op;
 		    __entry->usage = usage;
 		    __entry->where = where;
net/rxrpc/call_accept.c +1 -1
index c5566bc4aaca..47cf24630c70 100644
@@ -101,7 +101,7 @@ static int rxrpc_service_prealloc_one(struct rxrpc_sock *rx,
 	call->flags |= (1 << RXRPC_CALL_IS_SERVICE);
 	call->state = RXRPC_CALL_SERVER_PREALLOC;
 
-	trace_rxrpc_call(call, rxrpc_call_new_service,
+	trace_rxrpc_call(call->debug_id, rxrpc_call_new_service,
 			 atomic_read(&call->usage),
 			 here, (const void *)user_call_ID);
 
net/rxrpc/call_object.c +17 -11
index 215f4d98baa0..f58e624490a9 100644
@@ -242,7 +242,8 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx,
 	}
 
 	call->tx_total_len = p->tx_total_len;
-	trace_rxrpc_call(call, rxrpc_call_new_client, atomic_read(&call->usage),
+	trace_rxrpc_call(call->debug_id, rxrpc_call_new_client,
+			 atomic_read(&call->usage),
 			 here, (const void *)p->user_call_ID);
 
 	/* We need to protect a partially set up call against the user as we
@@ -292,8 +293,8 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx,
 	if (ret < 0)
 		goto error;
 
-	trace_rxrpc_call(call, rxrpc_call_connected, atomic_read(&call->usage),
-			 here, NULL);
+	trace_rxrpc_call(call->debug_id, rxrpc_call_connected,
+			 atomic_read(&call->usage), here, NULL);
 
 	rxrpc_start_call_timer(call);
 
@@ -315,8 +316,8 @@ struct rxrpc_call *rxrpc_new_client_call(struct rxrpc_sock *rx,
 error:
 	__rxrpc_set_call_completion(call, RXRPC_CALL_LOCAL_ERROR,
 				    RX_CALL_DEAD, ret);
-	trace_rxrpc_call(call, rxrpc_call_error, atomic_read(&call->usage),
-			 here, ERR_PTR(ret));
+	trace_rxrpc_call(call->debug_id, rxrpc_call_error,
+			 atomic_read(&call->usage), here, ERR_PTR(ret));
 	rxrpc_release_call(rx, call);
 	mutex_unlock(&call->user_mutex);
 	rxrpc_put_call(call, rxrpc_call_put);
@@ -420,7 +421,8 @@ bool rxrpc_queue_call(struct rxrpc_call *call)
 	if (n == 0)
 		return false;
 	if (rxrpc_queue_work(&call->processor))
-		trace_rxrpc_call(call, rxrpc_call_queued, n + 1, here, NULL);
+		trace_rxrpc_call(call->debug_id, rxrpc_call_queued, n + 1,
+				 here, NULL);
 	else
 		rxrpc_put_call(call, rxrpc_call_put_noqueue);
 	return true;
@@ -435,7 +437,8 @@ bool __rxrpc_queue_call(struct rxrpc_call *call)
 	int n = atomic_read(&call->usage);
 	ASSERTCMP(n, >=, 1);
 	if (rxrpc_queue_work(&call->processor))
-		trace_rxrpc_call(call, rxrpc_call_queued_ref, n, here, NULL);
+		trace_rxrpc_call(call->debug_id, rxrpc_call_queued_ref, n,
+				 here, NULL);
 	else
 		rxrpc_put_call(call, rxrpc_call_put_noqueue);
 	return true;
@@ -450,7 +453,8 @@ void rxrpc_see_call(struct rxrpc_call *call)
 	if (call) {
 		int n = atomic_read(&call->usage);
 
-		trace_rxrpc_call(call, rxrpc_call_seen, n, here, NULL);
+		trace_rxrpc_call(call->debug_id, rxrpc_call_seen, n,
+				 here, NULL);
 	}
 }
 
@@ -462,7 +466,7 @@ void rxrpc_get_call(struct rxrpc_call *call, enum rxrpc_call_trace op)
 	const void *here = __builtin_return_address(0);
 	int n = atomic_inc_return(&call->usage);
 
-	trace_rxrpc_call(call, op, n, here, NULL);
+	trace_rxrpc_call(call->debug_id, op, n, here, NULL);
 }
 
 /*
@@ -477,7 +481,8 @@ void rxrpc_release_call(struct rxrpc_sock *rx, struct rxrpc_call *call)
 
 	_enter("{%d,%d}", call->debug_id, atomic_read(&call->usage));
 
-	trace_rxrpc_call(call, rxrpc_call_release, atomic_read(&call->usage),
+	trace_rxrpc_call(call->debug_id, rxrpc_call_release,
+			 atomic_read(&call->usage),
 			 here, (const void *)call->flags);
 
 	ASSERTCMP(call->state, ==, RXRPC_CALL_COMPLETE);
@@ -625,12 +630,13 @@ void rxrpc_put_call(struct rxrpc_call *call, enum rxrpc_call_trace op)
 {
 	struct rxrpc_net *rxnet = call->rxnet;
 	const void *here = __builtin_return_address(0);
+	unsigned int debug_id = call->debug_id;
 	int n;
 
 	ASSERT(call != NULL);
 
 	n = atomic_dec_return(&call->usage);
-	trace_rxrpc_call(call, op, n, here, NULL);
+	trace_rxrpc_call(debug_id, op, n, here, NULL);
 	ASSERTCMP(n, >=, 0);
 	if (n == 0) {
 		_debug("call %d dead", call->debug_id);
Signature · 1 line
2.20.1
author maintainer
#319 lore raw

PATCH 578/671 mac80211: accept deauth frames in IBSS mode

In reply to Sasha Levin

From: Johannes Berg <[email protected]>

[ Upstream commit 95697f9907bfe3eab0ef20265a766b22e27dde64 ]

We can process deauth frames and all, but we drop them very early in the RX path today - this could never have worked.

Fixes 2cc59e784b54 ("mac80211: reply to AUTH with DEAUTH if sta allocation fails in IBSS")
---
 net/mac80211/rx.c | 11 ++++++++++-
 1 file changed, 10 insertions(+), 1 deletion(-)
net/mac80211/rx.c +10 -1
index b12f23c996f4..02d0b22d0114 100644
@@ -3391,9 +3391,18 @@ ieee80211_rx_h_mgmt(struct ieee80211_rx_data *rx)
 	case cpu_to_le16(IEEE80211_STYPE_PROBE_RESP):
 		/* process for all: mesh, mlme, ibss */
 		break;
+	case cpu_to_le16(IEEE80211_STYPE_DEAUTH):
+		if (is_multicast_ether_addr(mgmt->da) &&
+		    !is_broadcast_ether_addr(mgmt->da))
+			return RX_DROP_MONITOR;
+
+		/* process only for station/IBSS */
+		if (sdata->vif.type != NL80211_IFTYPE_STATION &&
+		    sdata->vif.type != NL80211_IFTYPE_ADHOC)
+			return RX_DROP_MONITOR;
+		break;
 	case cpu_to_le16(IEEE80211_STYPE_ASSOC_RESP):
 	case cpu_to_le16(IEEE80211_STYPE_REASSOC_RESP):
-	case cpu_to_le16(IEEE80211_STYPE_DEAUTH):
 	case cpu_to_le16(IEEE80211_STYPE_DISASSOC):
 		if (is_multicast_ether_addr(mgmt->da) &&
 		    !is_broadcast_ether_addr(mgmt->da))
Signature · 1 line
2.20.1
author maintainer
#320 lore raw

PATCH 579/671 llc: fix another potential sk_buff leak in llc_ui_sendmsg()

In reply to Sasha Levin

From: Eric Biggers <[email protected]>

[ Upstream commit fc8d5db10cbe1338a52ebc74e7feab9276721774 ]

All callers of llc_conn_state_process() except llc_build_and_send_pkt() (via llc_ui_sendmsg() -> llc_ui_send_data()) assume that it always consumes a reference to the skb. Fix this caller to do the same.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 net/llc/af_llc.c   | 34 ++++++++++++++++++++--------------
 net/llc/llc_conn.c |  2 ++
 net/llc/llc_if.c   | 12 ++++++++----
 3 files changed, 30 insertions(+), 18 deletions(-)
net/llc/af_llc.c +20 -14
index b99e73a7e7e0..ce841d59bc72 100644
@@ -113,22 +113,26 @@ static inline u8 llc_ui_header_len(struct sock *sk, struct sockaddr_llc *addr)
  *
  *	Send data via reliable llc2 connection.
  *	Returns 0 upon success, non-zero if action did not succeed.
+ *
+ *	This function always consumes a reference to the skb.
  */
 static int llc_ui_send_data(struct sock* sk, struct sk_buff *skb, int noblock)
 {
 	struct llc_sock* llc = llc_sk(sk);
-	int rc = 0;
 
 	if (unlikely(llc_data_accept_state(llc->state) ||
 		     llc->remote_busy_flag ||
 		     llc->p_flag)) {
 		long timeout = sock_sndtimeo(sk, noblock);
+		int rc;
 
 		rc = llc_ui_wait_for_busy_core(sk, timeout);
+		if (rc) {
+			kfree_skb(skb);
+			return rc;
+		}
 	}
-	if (unlikely(!rc))
-		rc = llc_build_and_send_pkt(sk, skb);
-	return rc;
+	return llc_build_and_send_pkt(sk, skb);
 }
 
 static void llc_ui_sk_init(struct socket *sock, struct sock *sk)
@@ -900,7 +904,7 @@ static int llc_ui_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	DECLARE_SOCKADDR(struct sockaddr_llc *, addr, msg->msg_name);
 	int flags = msg->msg_flags;
 	int noblock = flags & MSG_DONTWAIT;
-	struct sk_buff *skb;
+	struct sk_buff *skb = NULL;
 	size_t size = 0;
 	int rc = -EINVAL, copied = 0, hdrlen;
 
@@ -909,10 +913,10 @@ static int llc_ui_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	lock_sock(sk);
 	if (addr) {
 		if (msg->msg_namelen < sizeof(*addr))
-			goto release;
+			goto out;
 	} else {
 		if (llc_ui_addr_null(&llc->addr))
-			goto release;
+			goto out;
 		addr = &llc->addr;
 	}
 	/* must bind connection to sap if user hasn't done it. */
@@ -920,7 +924,7 @@ static int llc_ui_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 		/* bind to sap with null dev, exclusive. */
 		rc = llc_ui_autobind(sock, addr);
 		if (rc)
-			goto release;
+			goto out;
 	}
 	hdrlen = llc->dev->hard_header_len + llc_ui_header_len(sk, addr);
 	size = hdrlen + len;
@@ -929,12 +933,12 @@ static int llc_ui_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	copied = size - hdrlen;
 	rc = -EINVAL;
 	if (copied < 0)
-		goto release;
+		goto out;
 	release_sock(sk);
 	skb = sock_alloc_send_skb(sk, size, noblock, &rc);
 	lock_sock(sk);
 	if (!skb)
-		goto release;
+		goto out;
 	skb->dev      = llc->dev;
 	skb->protocol = llc_proto_type(addr->sllc_arphrd);
 	skb_reserve(skb, hdrlen);
@@ -944,29 +948,31 @@ static int llc_ui_sendmsg(struct socket *sock, struct msghdr *msg, size_t len)
 	if (sk->sk_type == SOCK_DGRAM || addr->sllc_ua) {
 		llc_build_and_send_ui_pkt(llc->sap, skb, addr->sllc_mac,
 					  addr->sllc_sap);
+		skb = NULL;
 		goto out;
 	}
 	if (addr->sllc_test) {
 		llc_build_and_send_test_pkt(llc->sap, skb, addr->sllc_mac,
 					    addr->sllc_sap);
+		skb = NULL;
 		goto out;
 	}
 	if (addr->sllc_xid) {
 		llc_build_and_send_xid_pkt(llc->sap, skb, addr->sllc_mac,
 					   addr->sllc_sap);
+		skb = NULL;
 		goto out;
 	}
 	rc = -ENOPROTOOPT;
 	if (!(sk->sk_type == SOCK_STREAM && !addr->sllc_ua))
 		goto out;
 	rc = llc_ui_send_data(sk, skb, noblock);
+	skb = NULL;
 out:
-	if (rc) {
-		kfree_skb(skb);
-release:
+	kfree_skb(skb);
+	if (rc)
 		dprintk("%s: failed sending from %02X to %02X: %d\n",
 			__func__, llc->laddr.lsap, llc->daddr.lsap, rc);
-	}
 	release_sock(sk);
 	return rc ? : copied;
 }
net/llc/llc_conn.c +2 -0
index ed2aca12460c..0b0c6f12153b 100644
@@ -55,6 +55,8 @@ int sysctl_llc2_busy_timeout = LLC2_BUSY_TIME * HZ;
  *	(executing it's actions and changing state), upper layer will be
  *	indicated or confirmed, if needed. Returns 0 for success, 1 for
  *	failure. The socket lock has to be held before calling this function.
+ *
+ *	This function always consumes a reference to the skb.
  */
 int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 {
net/llc/llc_if.c +8 -4
index 8db03c2d5440..ad6547736c21 100644
@@ -38,6 +38,8 @@
  *	closed and -EBUSY when sending data is not permitted in this state or
  *	LLC has send an I pdu with p bit set to 1 and is waiting for it's
  *	response.
+ *
+ *	This function always consumes a reference to the skb.
  */
 int llc_build_and_send_pkt(struct sock *sk, struct sk_buff *skb)
 {
@@ -46,20 +48,22 @@ int llc_build_and_send_pkt(struct sock *sk, struct sk_buff *skb)
 	struct llc_sock *llc = llc_sk(sk);
 
 	if (unlikely(llc->state == LLC_CONN_STATE_ADM))
-		goto out;
+		goto out_free;
 	rc = -EBUSY;
 	if (unlikely(llc_data_accept_state(llc->state) || /* data_conn_refuse */
 		     llc->p_flag)) {
 		llc->failed_data_req = 1;
-		goto out;
+		goto out_free;
 	}
 	ev = llc_conn_ev(skb);
 	ev->type      = LLC_CONN_EV_TYPE_PRIM;
 	ev->prim      = LLC_DATA_PRIM;
 	ev->prim_type = LLC_PRIM_TYPE_REQ;
 	skb->dev      = llc->dev;
-	rc = llc_conn_state_process(sk, skb);
-out:
+	return llc_conn_state_process(sk, skb);
+
+out_free:
+	kfree_skb(skb);
 	return rc;
 }
 
Signature · 1 line
2.20.1
author maintainer
#321 lore raw

PATCH 580/671 llc: fix sk_buff refcounting in llc_conn_state_process()

In reply to Sasha Levin

From: Eric Biggers <[email protected]>

[ Upstream commit 36453c852816f19947ca482a595dffdd2efa4965 ]

If llc_conn_state_process() sees that llc_conn_service() put the skb on a list, it will drop one fewer references to it. This is wrong because the current behavior is that llc_conn_service() never consumes a reference to the skb.

The code also makes the number of skb references being dropped conditional on which of ind_prim and cfm_prim are nonzero, yet neither of these affects how many references are *acquired*. So there is extra code that tries to fix this up by sometimes taking another reference.

Remove the unnecessary/broken refcounting logic and instead just add an skb_get() before the only two places where an extra reference is actually consumed.

Fixes 1da177e4c3f4 ("Linux-2.6.12-rc2")
---
 net/llc/llc_conn.c | 33 ++++++---------------------------
 1 file changed, 6 insertions(+), 27 deletions(-)
net/llc/llc_conn.c +6 -27
index 0b0c6f12153b..a79b739eb223 100644
@@ -64,12 +64,6 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 	struct llc_sock *llc = llc_sk(skb->sk);
 	struct llc_conn_state_ev *ev = llc_conn_ev(skb);
 
-	/*
-	 * We have to hold the skb, because llc_conn_service will kfree it in
-	 * the sending path and we need to look at the skb->cb, where we encode
-	 * llc_conn_state_ev.
-	 */
-	skb_get(skb);
 	ev->ind_prim = ev->cfm_prim = 0;
 	/*
 	 * Send event to state machine
@@ -77,21 +71,12 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 	rc = llc_conn_service(skb->sk, skb);
 	if (unlikely(rc != 0)) {
 		printk(KERN_ERR "%s: llc_conn_service failed\n", __func__);
-		goto out_kfree_skb;
-	}
-
-	if (unlikely(!ev->ind_prim && !ev->cfm_prim)) {
-		/* indicate or confirm not required */
-		if (!skb->next)
-			goto out_kfree_skb;
 		goto out_skb_put;
 	}
 
-	if (unlikely(ev->ind_prim && ev->cfm_prim)) /* Paranoia */
-		skb_get(skb);
-
 	switch (ev->ind_prim) {
 	case LLC_DATA_PRIM:
+		skb_get(skb);
 		llc_save_primitive(sk, skb, LLC_DATA_PRIM);
 		if (unlikely(sock_queue_rcv_skb(sk, skb))) {
 			/*
@@ -108,6 +93,7 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 		 * skb->sk pointing to the newly created struct sock in
 		 * llc_conn_handler. -acme
 		 */
+		skb_get(skb);
 		skb_queue_tail(&sk->sk_receive_queue, skb);
 		sk->sk_state_change(sk);
 		break;
@@ -123,7 +109,6 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 				sk->sk_state_change(sk);
 			}
 		}
-		kfree_skb(skb);
 		sock_put(sk);
 		break;
 	case LLC_RESET_PRIM:
@@ -132,14 +117,11 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 		 * RESET is not being notified to upper layers for now
 		 */
 		printk(KERN_INFO "%s: received a reset ind!\n", __func__);
-		kfree_skb(skb);
 		break;
 	default:
-		if (ev->ind_prim) {
+		if (ev->ind_prim)
 			printk(KERN_INFO "%s: received unknown %d prim!\n",
 				__func__, ev->ind_prim);
-			kfree_skb(skb);
-		}
 		/* No indication */
 		break;
 	}
@@ -181,15 +163,12 @@ int llc_conn_state_process(struct sock *sk, struct sk_buff *skb)
 		printk(KERN_INFO "%s: received a reset conf!\n", __func__);
 		break;
 	default:
-		if (ev->cfm_prim) {
+		if (ev->cfm_prim)
 			printk(KERN_INFO "%s: received unknown %d prim!\n",
 					__func__, ev->cfm_prim);
-			break;
-		}
-		goto out_skb_put; /* No confirmation */
+		/* No confirmation */
+		break;
 	}
-out_kfree_skb:
-	kfree_skb(skb);
 out_skb_put:
 	kfree_skb(skb);
 	return rc;
Signature · 1 line
2.20.1
author maintainer
#322 lore raw

PATCH 581/671 ip6erspan: remove the incorrect mtu limit for ip6erspan

In reply to Sasha Levin

From: Haishuang Yan <[email protected]>

[ Upstream commit 4123f637a5129470ff9d3cb00a5a4e213f2e15cc ]

ip6erspan driver calls ether_setup(), after commit 61e84623ace3 ("net: centralize net_device min/max MTU checking"), the range of mtu is [min_mtu, max_mtu], which is [68, 1500] by default.

It causes the dev mtu of the erspan device to not be greater than 1500, this limit value is not correct for ip6erspan tap device.

Fixes 61e84623ace3 ("net: centralize net_device min/max MTU checking")
---
 net/ipv6/ip6_gre.c | 1 +
 1 file changed, 1 insertion(+)
net/ipv6/ip6_gre.c +1 -0
index b3515a4f1303..1f2d0022ba6f 100644
@@ -2218,6 +2218,7 @@ static void ip6erspan_tap_setup(struct net_device *dev)
 {
 	ether_setup(dev);
 
+	dev->max_mtu = 0;
 	dev->netdev_ops = &ip6erspan_netdev_ops;
 	dev->needs_free_netdev = true;
 	dev->priv_destructor = ip6gre_dev_free;
Signature · 1 line
2.20.1
author maintainer
#323 lore raw

PATCH 582/671 iwlwifi: pcie: fix memory leaks in iwl_pcie_ctxt_info_gen3_init

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit 0f4f199443faca715523b0659aa536251d8b978f ]

In iwl_pcie_ctxt_info_gen3_init there are cases that the allocated dma memory is leaked in case of error.

DMA memories prph_scratch, prph_info, and ctxt_info_gen3 are allocated and initialized to be later assigned to trans_pcie. But in any error case before such assignment the allocated memories should be released.

First of such error cases happens when iwl_pcie_init_fw_sec fails. Current implementation correctly releases prph_scratch. But in two sunsequent error cases where dma_alloc_coherent may fail, such releases are missing.

This commit adds release for prph_scratch when allocation for prph_info fails, and adds releases for prph_scratch and prph_info when allocation for ctxt_info_gen3 fails.

Fixes 2ee824026288 ("iwlwifi: pcie: support context information for 22560 devices")
---
 .../intel/iwlwifi/pcie/ctxt-info-gen3.c       | 36 +++++++++++++------
 1 file changed, 25 insertions(+), 11 deletions(-)
drivers/net/wireless/intel/iwlwifi/pcie/ctxt-info-gen3.c +25 -11
index 64d976d872b8..6783b20d9681 100644
@@ -102,13 +102,9 @@ int iwl_pcie_ctxt_info_gen3_init(struct iwl_trans *trans,
 
 	/* allocate ucode sections in dram and set addresses */
 	ret = iwl_pcie_init_fw_sec(trans, fw, &prph_scratch->dram);
-	if (ret) {
-		dma_free_coherent(trans->dev,
-				  sizeof(*prph_scratch),
-				  prph_scratch,
-				  trans_pcie->prph_scratch_dma_addr);
-		return ret;
-	}
+	if (ret)
+		goto err_free_prph_scratch;
+
 
 	/* Allocate prph information
 	 * currently we don't assign to the prph info anything, but it would get
@@ -116,16 +112,20 @@ int iwl_pcie_ctxt_info_gen3_init(struct iwl_trans *trans,
 	prph_info = dma_alloc_coherent(trans->dev, sizeof(*prph_info),
 				       &trans_pcie->prph_info_dma_addr,
 				       GFP_KERNEL);
-	if (!prph_info)
-		return -ENOMEM;
+	if (!prph_info) {
+		ret = -ENOMEM;
+		goto err_free_prph_scratch;
+	}
 
 	/* Allocate context info */
 	ctxt_info_gen3 = dma_alloc_coherent(trans->dev,
 					    sizeof(*ctxt_info_gen3),
 					    &trans_pcie->ctxt_info_dma_addr,
 					    GFP_KERNEL);
-	if (!ctxt_info_gen3)
-		return -ENOMEM;
+	if (!ctxt_info_gen3) {
+		ret = -ENOMEM;
+		goto err_free_prph_info;
+	}
 
 	ctxt_info_gen3->prph_info_base_addr =
 		cpu_to_le64(trans_pcie->prph_info_dma_addr);
@@ -176,6 +176,20 @@ int iwl_pcie_ctxt_info_gen3_init(struct iwl_trans *trans,
 	iwl_set_bit(trans, CSR_GP_CNTRL, CSR_AUTO_FUNC_INIT);
 
 	return 0;
+
+err_free_prph_info:
+	dma_free_coherent(trans->dev,
+			  sizeof(*prph_info),
+			prph_info,
+			trans_pcie->prph_info_dma_addr);
+
+err_free_prph_scratch:
+	dma_free_coherent(trans->dev,
+			  sizeof(*prph_scratch),
+			prph_scratch,
+			trans_pcie->prph_scratch_dma_addr);
+	return ret;
+
 }
 
 void iwl_pcie_ctxt_info_gen3_free(struct iwl_trans *trans)
Signature · 1 line
2.20.1
author maintainer
#324 lore raw

PATCH 583/671 net: stmmac: fix length of PTP clock's name string

In reply to Sasha Levin

From: Antonio Borneo <[email protected]>

[ Upstream commit 5da202c88f8c355ad79bc2e8eb582e6d433060e7 ]

The field "name" in struct ptp_clock_info has a fixed size of 16 chars and is used as zero terminated string by clock_name_show() in drivers/ptp/ptp_sysfs.c
The current initialization value requires 17 chars to fit also the null termination, and this causes overflow to the next bytes in the struct when the string is read as null terminated:

	hexdump -C /sys/class/ptp/ptp0/clock_name
	00000000  73 74 6d 6d 61 63 5f 70  74 70 5f 63 6c 6f 63 6b  |stmmac_ptp_clock|
	00000010  a0 ac b9 03 0a                                    |.....|

where the extra 4 bytes (excluding the newline) after the string represent the integer 0x03b9aca0 = 62500000 assigned to the field "max_adj" that follows "name" in the same struct.

There is no strict requirement for the "name" content and in the comment in ptp_clock_kernel.h it's reported it should just be 'A short "friendly name" to identify the clock'.
Replace it with "stmmac ptp".

Fixes 92ba6888510c ("stmmac: add the support for PTP hw clock driver")
---
 drivers/net/ethernet/stmicro/stmmac/stmmac_ptp.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/stmicro/stmmac/stmmac_ptp.c +1 -1
index cc60b3fb0892..8f8b8f381ffd 100644
@@ -174,7 +174,7 @@ static int stmmac_enable(struct ptp_clock_info *ptp,
 /* structure describing a PTP hardware clock */
 static struct ptp_clock_info stmmac_ptp_clock_ops = {
 	.owner = THIS_MODULE,
-	.name = "stmmac_ptp_clock",
+	.name = "stmmac ptp",
 	.max_adj = 62500000,
 	.n_alarm = 0,
 	.n_ext_ts = 0,
Signature · 1 line
2.20.1
author maintainer
#325 lore raw

PATCH 584/671 net: stmmac: fix disabling flexible PPS output

In reply to Sasha Levin

From: Antonio Borneo <[email protected]>

[ Upstream commit 520cf6002147281d1e7b522bb338416b623dcb93 ]

Accordingly to Synopsys documentation [1] and [2], when bit PPSEN0 in register MAC_PPS_CONTROL is set it selects the functionality command in the same register, otherwise selects the functionality control.
Command functionality is required to either enable (command 0x2) and disable (command 0x5) the flexible PPS output, but the bit PPSEN0 is currently set only for enabling.

Set the bit PPSEN0 to properly disable flexible PPS output.

Tested on STM32MP15x, based on dwmac 4.10a.

[1] DWC Ethernet QoS Databook 4.10a October 2014
[2] DWC Ethernet QoS Databook 5.00a September 2017

Fixes 9a8a02c9d46d ("net: stmmac: Add Flexible PPS support")
---
 drivers/net/ethernet/stmicro/stmmac/dwmac5.c | 1 +
 1 file changed, 1 insertion(+)
drivers/net/ethernet/stmicro/stmmac/dwmac5.c +1 -0
index 3f4f3132e16b..e436fa160c7d 100644
@@ -515,6 +515,7 @@ int dwmac5_flex_pps_config(void __iomem *ioaddr, int index,
 
 	if (!enable) {
 		val |= PPSCMDx(index, 0x5);
+		val |= PPSEN0;
 		writel(val, ioaddr + MAC_PPS_CONTROL);
 		return 0;
 	}
Signature · 7 lines
2.20.1


_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
author maintainer
#326 lore raw

PATCH 585/671 sctp: add chunks to sk_backlog when the newsk sk_socket is not set

In reply to Sasha Levin

From: Xin Long <[email protected]>

[ Upstream commit 819be8108fded0b9e710bbbf81193e52f7bab2f7 ]

This patch is to fix a NULL-ptr deref in selinux_socket_connect_helper:

[...] kasan: GPF could be caused by NULL-ptr deref or user memory access [...] RIP: 0010:selinux_socket_connect_helper+0x94/0x460 [...] Call Trace:

  [...]  selinux_sctp_bind_connect+0x16a/0x1d0
  [...]  security_sctp_bind_connect+0x58/0x90
  [...]  sctp_process_asconf+0xa52/0xfd0 [sctp]
  [...]  sctp_sf_do_asconf+0x785/0x980 [sctp]
  [...]  sctp_do_sm+0x175/0x5a0 [sctp]
  [...]  sctp_assoc_bh_rcv+0x285/0x5b0 [sctp]
  [...]  sctp_backlog_rcv+0x482/0x910 [sctp]
  [...]  __release_sock+0x11e/0x310
  [...]  release_sock+0x4f/0x180
  [...]  sctp_accept+0x3f9/0x5a0 [sctp]
  [...]  inet_accept+0xe7/0x720

It was caused by that the 'newsk' sk_socket was not set before going to security sctp hook when processing asconf chunk with SCTP_PARAM_ADD_IP or SCTP_PARAM_SET_PRIMARY:

inet_accept()->

    sctp_accept():
      lock_sock():
          lock listening 'sk'
                                          do_softirq():
                                            sctp_rcv():  <-- [1]
                                                asconf chunk arrives and
                                                enqueued in 'sk' backlog
      sctp_sock_migrate():
          set asoc's sk to 'newsk'
      release_sock():
          sctp_backlog_rcv():
            lock 'newsk'
            sctp_process_asconf()  <-- [2]
            unlock 'newsk'
    sock_graft():
        set sk_socket  <-- [3]

As it shows, at [1] the asconf chunk would be put into the listening 'sk' backlog, as accept() was holding its sock lock. Then at [2] asconf would get processed with 'newsk' as asoc's sk had been set to 'newsk'. However, 'newsk' sk_socket is not set until [3], while selinux_sctp_bind_connect() would deref it, then kernel crashed.

Here to fix it by adding the chunk to sk_backlog until newsk sk_socket is set when .accept() is done.

Note that sk->sk_socket can be NULL when the sock is closed, so SOCK_DEAD flag is also needed to check in sctp_newsk_ready().

Thanks to Ondrej for reviewing the code.

---
 include/net/sctp/sctp.h |  5 +++++
 net/sctp/input.c        | 12 +++++++++---
 2 files changed, 14 insertions(+), 3 deletions(-)
include/net/sctp/sctp.h +5 -0
index 2abbc15824af..2c6570e6fcfe 100644
@@ -625,4 +625,9 @@ static inline __u32 sctp_min_frag_point(struct sctp_sock *sp, __u16 datasize)
 	return sctp_mtu_payload(sp, SCTP_DEFAULT_MINSEGMENT, datasize);
 }
 
+static inline bool sctp_newsk_ready(const struct sock *sk)
+{
+	return sock_flag(sk, SOCK_DEAD) || sk->sk_socket;
+}
+
 #endif /* __net_sctp_h__ */
net/sctp/input.c +9 -3
index bfe29158afcc..f64d882c8698 100644
@@ -255,7 +255,7 @@ int sctp_rcv(struct sk_buff *skb)
 		bh_lock_sock(sk);
 	}
 
-	if (sock_owned_by_user(sk)) {
+	if (sock_owned_by_user(sk) || !sctp_newsk_ready(sk)) {
 		if (sctp_add_backlog(sk, skb)) {
 			bh_unlock_sock(sk);
 			sctp_chunk_free(chunk);
@@ -333,7 +333,7 @@ int sctp_backlog_rcv(struct sock *sk, struct sk_buff *skb)
 		local_bh_disable();
 		bh_lock_sock(sk);
 
-		if (sock_owned_by_user(sk)) {
+		if (sock_owned_by_user(sk) || !sctp_newsk_ready(sk)) {
 			if (sk_add_backlog(sk, skb, sk->sk_rcvbuf))
 				sctp_chunk_free(chunk);
 			else
@@ -348,7 +348,13 @@ int sctp_backlog_rcv(struct sock *sk, struct sk_buff *skb)
 		if (backloged)
 			return 0;
 	} else {
-		sctp_inq_push(inqueue, chunk);
+		if (!sctp_newsk_ready(sk)) {
+			if (!sk_add_backlog(sk, skb, sk->sk_rcvbuf))
+				return 0;
+			sctp_chunk_free(chunk);
+		} else {
+			sctp_inq_push(inqueue, chunk);
+		}
 	}
 
 done:
Signature · 1 line
2.20.1
author maintainer
#327 lore raw

PATCH 586/671 s390/qeth: Fix error handling during VNICC initialization

In reply to Sasha Levin

From: Alexandra Winter <[email protected]>

[ Upstream commit b528965bcc827dad32a8d21745feaacfc76c9703 ]

Smatch discovered the use of uninitialized variable sup_cmds in error paths.

Fixes caa1f0b10d18 ("s390/qeth: add VNICC enable/disable support")
---
 drivers/s390/net/qeth_l2_main.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)
drivers/s390/net/qeth_l2_main.c +7 -4
index 95669d47c389..cf3b15ebcf75 100644
@@ -2331,10 +2331,10 @@ static bool qeth_l2_vnicc_recover_char(struct qeth_card *card, u32 vnicc,
 static void qeth_l2_vnicc_init(struct qeth_card *card)
 {
 	u32 *timeout = &card->options.vnicc.learning_timeout;
+	bool enable, error = false;
 	unsigned int chars_len, i;
 	unsigned long chars_tmp;
 	u32 sup_cmds, vnicc;
-	bool enable, error;
 
 	QETH_CARD_TEXT(card, 2, "vniccini");
 	/* reset rx_bcast */
@@ -2355,7 +2355,10 @@ static void qeth_l2_vnicc_init(struct qeth_card *card)
 	chars_len = sizeof(card->options.vnicc.sup_chars) * BITS_PER_BYTE;
 	for_each_set_bit(i, &chars_tmp, chars_len) {
 		vnicc = BIT(i);
-		qeth_l2_vnicc_query_cmds(card, vnicc, &sup_cmds);
+		if (qeth_l2_vnicc_query_cmds(card, vnicc, &sup_cmds)) {
+			sup_cmds = 0;
+			error = true;
+		}
 		if (!(sup_cmds & IPA_VNICC_SET_TIMEOUT) ||
 		    !(sup_cmds & IPA_VNICC_GET_TIMEOUT))
 			card->options.vnicc.getset_timeout_sup &= ~vnicc;
@@ -2364,8 +2367,8 @@ static void qeth_l2_vnicc_init(struct qeth_card *card)
 			card->options.vnicc.set_char_sup &= ~vnicc;
 	}
 	/* enforce assumed default values and recover settings, if changed  */
-	error = qeth_l2_vnicc_recover_timeout(card, QETH_VNICC_LEARNING,
-					      timeout);
+	error |= qeth_l2_vnicc_recover_timeout(card, QETH_VNICC_LEARNING,
+					       timeout);
 	chars_tmp = card->options.vnicc.wanted_chars ^ QETH_VNICC_DEFAULT;
 	chars_tmp |= QETH_VNICC_BRIDGE_INVISIBLE;
 	chars_len = sizeof(card->options.vnicc.wanted_chars) * BITS_PER_BYTE;
Signature · 1 line
2.20.1
author maintainer
#328 lore raw

PATCH 587/671 s390/qeth: Fix initialization of vnicc cmd masks during set online

In reply to Sasha Levin

From: Alexandra Winter <[email protected]>

[ Upstream commit be40a86c319706f90caca144343c64743c32b953 ]

Without this patch, a command bit in the supported commands mask is only ever set to unsupported during set online. If a command is ever marked as unsupported (e.g. because of error during qeth_l2_vnicc_query_cmds), subsequent successful initialization (offline/online) would not bring it back.

Fixes caa1f0b10d18 ("s390/qeth: add VNICC enable/disable support")
---
 drivers/s390/net/qeth_l2_main.c | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)
drivers/s390/net/qeth_l2_main.c +8 -4
index cf3b15ebcf75..f86465da2422 100644
@@ -2359,11 +2359,15 @@ static void qeth_l2_vnicc_init(struct qeth_card *card)
 			sup_cmds = 0;
 			error = true;
 		}
-		if (!(sup_cmds & IPA_VNICC_SET_TIMEOUT) ||
-		    !(sup_cmds & IPA_VNICC_GET_TIMEOUT))
+		if ((sup_cmds & IPA_VNICC_SET_TIMEOUT) &&
+		    (sup_cmds & IPA_VNICC_GET_TIMEOUT))
+			card->options.vnicc.getset_timeout_sup |= vnicc;
+		else
 			card->options.vnicc.getset_timeout_sup &= ~vnicc;
-		if (!(sup_cmds & IPA_VNICC_ENABLE) ||
-		    !(sup_cmds & IPA_VNICC_DISABLE))
+		if ((sup_cmds & IPA_VNICC_ENABLE) &&
+		    (sup_cmds & IPA_VNICC_DISABLE))
+			card->options.vnicc.set_char_sup |= vnicc;
+		else
 			card->options.vnicc.set_char_sup &= ~vnicc;
 	}
 	/* enforce assumed default values and recover settings, if changed  */
Signature · 1 line
2.20.1
author maintainer
#329 lore raw

PATCH 588/671 act_mirred: Fix mirred_init_module error handling

In reply to Sasha Levin

From: YueHaibing <[email protected]>

[ Upstream commit 11c9a7d38af524217efb7a176ad322b97ac2f163 ]

If tcf_register_action failed, mirred_device_notifier should be unregistered.

Fixes 3b87956ea645 ("net sched: fix race in mirred device removal")
---
 net/sched/act_mirred.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
net/sched/act_mirred.c +5 -1
index 399e3beae6cf..a30c17a28281 100644
@@ -445,7 +445,11 @@ static int __init mirred_init_module(void)
 		return err;
 
 	pr_info("Mirror/redirect action on\n");
-	return tcf_register_action(&act_mirred_ops, &mirred_net_ops);
+	err = tcf_register_action(&act_mirred_ops, &mirred_net_ops);
+	if (err)
+		unregister_netdevice_notifier(&mirred_device_notifier);
+
+	return err;
 }
 
 static void __exit mirred_cleanup_module(void)
Signature · 1 line
2.20.1
author maintainer
#330 lore raw

PATCH 589/671 net: avoid possible false sharing in sk_leave_memory_pressure()

In reply to Sasha Levin

From: Eric Dumazet <[email protected]>

[ Upstream commit 503978aca46124cd714703e180b9c8292ba50ba7 ]

As mentioned in https://github.com/google/ktsan/wiki/READ_ONCE-and-WRITE_ONCE#it-may-improve-performance
a C compiler can legally transform :

if (memory_pressure && *memory_pressure)

        *memory_pressure = 0;

to :

if (memory_pressure)

        *memory_pressure = 0;
Fixes 0604475119de ("tcp: add TCPMemoryPressuresChrono counter")
Fixes 180d8cd942ce ("foundations of per-cgroup memory pressure controlling.")
Fixes 3ab224be6d69 ("[NET] CORE: Introducing new memory accounting interface.")
---
 net/core/sock.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
net/core/sock.c +2 -2
index bbde5f6a7dc9..b9ec14f2c729 100644
@@ -2179,8 +2179,8 @@ static void sk_leave_memory_pressure(struct sock *sk)
 	} else {
 		unsigned long *memory_pressure = sk->sk_prot->memory_pressure;
 
-		if (memory_pressure && *memory_pressure)
-			*memory_pressure = 0;
+		if (memory_pressure && READ_ONCE(*memory_pressure))
+			WRITE_ONCE(*memory_pressure, 0);
 	}
 }
 
Signature · 1 line
2.20.1
author maintainer
#331 lore raw

PATCH 590/671 net: add {READ|WRITE}_ONCE() annotations on ->rskq_accept_head

In reply to Sasha Levin

From: Eric Dumazet <[email protected]>

[ Upstream commit 60b173ca3d1cd1782bd0096dc17298ec242f6fb1 ]

reqsk_queue_empty() is called from inet_csk_listen_poll() while other cpus might write ->rskq_accept_head value.

Use {READ|WRITE}_ONCE() to avoid compiler tricks
and potential KCSAN splats.

Fixes fff1f3001cc5 ("tcp: add a spinlock to protect struct request_sock_queue")
---
 drivers/xen/pvcalls-back.c      | 2 +-
 include/net/request_sock.h      | 4 ++--
 net/ipv4/inet_connection_sock.c | 2 +-
 3 files changed, 4 insertions(+), 4 deletions(-)
drivers/xen/pvcalls-back.c +1 -1
index d4ea33581ac2..b3fbfed28682 100644
@@ -784,7 +784,7 @@ static int pvcalls_back_poll(struct xenbus_device *dev,
 	mappass->reqcopy = *req;
 	icsk = inet_csk(mappass->sock->sk);
 	queue = &icsk->icsk_accept_queue;
-	data = queue->rskq_accept_head != NULL;
+	data = READ_ONCE(queue->rskq_accept_head) != NULL;
 	if (data) {
 		mappass->reqcopy.cmd = 0;
 		ret = 0;
include/net/request_sock.h +2 -2
index 347015515a7d..1653435f18f5 100644
@@ -183,7 +183,7 @@ void reqsk_fastopen_remove(struct sock *sk, struct request_sock *req,
 
 static inline bool reqsk_queue_empty(const struct request_sock_queue *queue)
 {
-	return queue->rskq_accept_head == NULL;
+	return READ_ONCE(queue->rskq_accept_head) == NULL;
 }
 
 static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue *queue,
@@ -195,7 +195,7 @@ static inline struct request_sock *reqsk_queue_remove(struct request_sock_queue
 	req = queue->rskq_accept_head;
 	if (req) {
 		sk_acceptq_removed(parent);
-		queue->rskq_accept_head = req->dl_next;
+		WRITE_ONCE(queue->rskq_accept_head, req->dl_next);
 		if (queue->rskq_accept_head == NULL)
 			queue->rskq_accept_tail = NULL;
 	}
net/ipv4/inet_connection_sock.c +1 -1
index 636a11c56cf5..79320858e719 100644
@@ -937,7 +937,7 @@ struct sock *inet_csk_reqsk_queue_add(struct sock *sk,
 		req->sk = child;
 		req->dl_next = NULL;
 		if (queue->rskq_accept_head == NULL)
-			queue->rskq_accept_head = req;
+			WRITE_ONCE(queue->rskq_accept_head, req);
 		else
 			queue->rskq_accept_tail->dl_next = req;
 		queue->rskq_accept_tail = req;
Signature · 1 line
2.20.1
author maintainer
#332 lore raw

PATCH 591/671 tcp: annotate lockless access to tcp_memory_pressure

In reply to Sasha Levin

From: Eric Dumazet <[email protected]>

[ Upstream commit 1f142c17d19a5618d5a633195a46f2c8be9bf232 ]

tcp_memory_pressure is read without holding any lock, and its value could be changed on other cpus.

Use READ_ONCE() to annotate these lockless reads.

The write side is already using atomic ops.

Fixes b8da51ebb1aa ("tcp: introduce tcp_under_memory_pressure()")
---
 include/net/tcp.h | 2 +-
 net/ipv4/tcp.c    | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)
include/net/tcp.h +1 -1
index ac4ffe8013d8..918bfd0d7d1f 100644
@@ -261,7 +261,7 @@ static inline bool tcp_under_memory_pressure(const struct sock *sk)
 	    mem_cgroup_under_socket_pressure(sk->sk_memcg))
 		return true;
 
-	return tcp_memory_pressure;
+	return READ_ONCE(tcp_memory_pressure);
 }
 /*
  * The next routines deal with comparing 32 bit unsigned ints
net/ipv4/tcp.c +2 -2
index 7561fa1bcc3e..3c181ca714d0 100644
@@ -325,7 +325,7 @@ void tcp_enter_memory_pressure(struct sock *sk)
 {
 	unsigned long val;
 
-	if (tcp_memory_pressure)
+	if (READ_ONCE(tcp_memory_pressure))
 		return;
 	val = jiffies;
 
@@ -340,7 +340,7 @@ void tcp_leave_memory_pressure(struct sock *sk)
 {
 	unsigned long val;
 
-	if (!tcp_memory_pressure)
+	if (!READ_ONCE(tcp_memory_pressure))
 		return;
 	val = xchg(&tcp_memory_pressure, 0);
 	if (val)
Signature · 1 line
2.20.1
author maintainer
#333 lore raw

PATCH 592/671 USB: usb-skeleton: fix use-after-free after driver unbind

In reply to Sasha Levin

From: Johan Hovold <[email protected]>

[ Upstream commit 6353001852776e7eeaab4da78922d4c6f2b076af ]

The driver failed to stop its read URB on disconnect, something which could lead to a use-after-free in the completion handler after driver unbind in case the character device has been closed.

Fixes e7389cc9a7ff ("USB: skel_read really sucks royally")
---
 drivers/usb/usb-skeleton.c | 1 +
 1 file changed, 1 insertion(+)
drivers/usb/usb-skeleton.c +1 -0
index e0cf11f798c5..2e99da45ce0b 100644
@@ -573,6 +573,7 @@ static void skel_disconnect(struct usb_interface *interface)
 	dev->disconnected = 1;
 	mutex_unlock(&dev->io_mutex);
 
+	usb_kill_urb(dev->bulk_in_urb);
 	usb_kill_anchored_urbs(&dev->submitted);
 
 	/* decrement our usage count */
Signature · 1 line
2.20.1
#334 lore raw

Re: [PATCH AUTOSEL 4.19 592/671] USB: usb-skeleton: fix use-after-free after driver unbind

Quoted 8 lines · Sasha Levin

From: Johan Hovold <[email protected]>

[ Upstream commit 6353001852776e7eeaab4da78922d4c6f2b076af ]

The driver failed to stop its read URB on disconnect, something which could lead to a use-after-free in the completion handler after driver unbind in case the character device has been closed.

Quoted 5 lines
Fixes e7389cc9a7ff ("USB: skel_read really sucks royally")

This one isn't needed in any stable tree. As we discussed before, the skeleton driver is only there for documentation purposes.

Johan

author maintainer
#335 lore raw

Re: [PATCH AUTOSEL 4.19 592/671] USB: usb-skeleton: fix use-after-free after driver unbind

Quoted 9 lines · Johan Hovold
Quoted 8 lines · Sasha Levin

From: Johan Hovold <[email protected]>

[ Upstream commit 6353001852776e7eeaab4da78922d4c6f2b076af ]

The driver failed to stop its read URB on disconnect, something which could lead to a use-after-free in the completion handler after driver unbind in case the character device has been closed.

Quoted 8 lines
Quoted 5 lines
Fixes e7389cc9a7ff ("USB: skel_read really sucks royally")

This one isn't needed in any stable tree. As we discussed before, the skeleton driver is only there for documentation purposes.

I'll drop this, but I'm curious: doesn't this mean that users will build on buggy example code?

Signature · 2 lines
Thanks,
Sasha
#336 lore raw

Re: [PATCH AUTOSEL 4.19 592/671] USB: usb-skeleton: fix use-after-free after driver unbind

Quoted 16 lines · Sasha Levin
Quoted 15 lines · Johan Hovold
Quoted 13 lines · Sasha Levin

From: Johan Hovold <[email protected]>

[ Upstream commit 6353001852776e7eeaab4da78922d4c6f2b076af ]

The driver failed to stop its read URB on disconnect, something which could lead to a use-after-free in the completion handler after driver unbind in case the character device has been closed.

Fixes e7389cc9a7ff ("USB: skel_read really sucks royally")
Quoted 5 lines
Quoted 2 lines

This one isn't needed in any stable tree. As we discussed before, the skeleton driver is only there for documentation purposes.

I'll drop this, but I'm curious: doesn't this mean that users will build on buggy example code?

They should always be grabbing the "latest" version of the file to work off of anyway. Given that there are very few new USB drivers anymore, I doubt this really matters much...

thanks,

greg k-h

author maintainer
#337 lore raw

PATCH 593/671 net/smc: receive returns without data

In reply to Sasha Levin

From: Karsten Graul <[email protected]>

[ Upstream commit 882dcfe5a1785c20f45820cbe6fec4b8b647c946 ]

smc_cdc_rxed_any_close_or_senddone() is used as an end condition for the receive loop. This conflicts with smc_cdc_msg_recv_action() which could run in parallel and set the bits checked by smc_cdc_rxed_any_close_or_senddone() before the receive is processed. In that case we could return from receive with no data, although data is available. The same applies to smc_rx_wait().
Fix this by checking for RCV_SHUTDOWN only, which is set in smc_cdc_msg_recv_action() after the receive was actually processed.

Fixes 952310ccf2d8 ("smc: receive data from RMBE")
---
 net/smc/smc_rx.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)
net/smc/smc_rx.c +1 -3
index bbcf0fe4ae10..1ee5fdbf8284 100644
@@ -212,8 +212,7 @@ int smc_rx_wait(struct smc_sock *smc, long *timeo,
 	rc = sk_wait_event(sk, timeo,
 			   sk->sk_err ||
 			   sk->sk_shutdown & RCV_SHUTDOWN ||
-			   fcrit(conn) ||
-			   smc_cdc_rxed_any_close_or_senddone(conn),
+			   fcrit(conn),
 			   &wait);
 	remove_wait_queue(sk_sleep(sk), &wait);
 	sk_clear_bit(SOCKWQ_ASYNC_WAITDATA, sk);
@@ -311,7 +310,6 @@ int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg,
 			smc_rx_update_cons(smc, 0);
 
 		if (sk->sk_shutdown & RCV_SHUTDOWN ||
-		    smc_cdc_rxed_any_close_or_senddone(conn) ||
 		    conn->local_tx_ctrl.conn_state_flags.peer_conn_abort)
 			break;
 
Signature · 1 line
2.20.1
author maintainer
#338 lore raw

PATCH 594/671 net/smc: receive pending data after RCV_SHUTDOWN

In reply to Sasha Levin

From: Karsten Graul <[email protected]>

[ Upstream commit 107529e31a87acd475ff6a0f82745821b8f70fec ]

smc_rx_recvmsg() first checks if data is available, and then if RCV_SHUTDOWN is set. There is a race when smc_cdc_msg_recv_action() runs in between these 2 checks, receives data and sets RCV_SHUTDOWN. In that case smc_rx_recvmsg() would return from receive without to process the available data.
Fix that with a final check for data available if RCV_SHUTDOWN is set. Move the check for data into a function and call it twice. And use the existing helper smc_rx_data_available().

Fixes 952310ccf2d8 ("smc: receive data from RMBE")
---
 net/smc/smc_rx.c | 25 ++++++++++++++++++++-----
 1 file changed, 20 insertions(+), 5 deletions(-)
net/smc/smc_rx.c +20 -5
index 1ee5fdbf8284..36340912df48 100644
@@ -262,6 +262,18 @@ static int smc_rx_recv_urg(struct smc_sock *smc, struct msghdr *msg, int len,
 	return -EAGAIN;
 }
 
+static bool smc_rx_recvmsg_data_available(struct smc_sock *smc)
+{
+	struct smc_connection *conn = &smc->conn;
+
+	if (smc_rx_data_available(conn))
+		return true;
+	else if (conn->urg_state == SMC_URG_VALID)
+		/* we received a single urgent Byte - skip */
+		smc_rx_update_cons(smc, 0);
+	return false;
+}
+
 /* smc_rx_recvmsg - receive data from RMBE
  * @msg:	copy data to receive buffer
  * @pipe:	copy data to pipe if set - indicates splice() call
@@ -303,15 +315,18 @@ int smc_rx_recvmsg(struct smc_sock *smc, struct msghdr *msg,
 		if (read_done >= target || (pipe && read_done))
 			break;
 
-		if (atomic_read(&conn->bytes_to_rcv))
+		if (smc_rx_recvmsg_data_available(smc))
 			goto copy;
-		else if (conn->urg_state == SMC_URG_VALID)
-			/* we received a single urgent Byte - skip */
-			smc_rx_update_cons(smc, 0);
 
 		if (sk->sk_shutdown & RCV_SHUTDOWN ||
-		    conn->local_tx_ctrl.conn_state_flags.peer_conn_abort)
+		    conn->local_tx_ctrl.conn_state_flags.peer_conn_abort) {
+			/* smc_cdc_msg_recv_action() could have run after
+			 * above smc_rx_recvmsg_data_available()
+			 */
+			if (smc_rx_recvmsg_data_available(smc))
+				goto copy;
 			break;
+		}
 
 		if (read_done) {
 			if (sk->sk_err ||
Signature · 1 line
2.20.1
author maintainer
#339 lore raw

PATCH 595/671 drm/msm/dsi: Implement reset correctly

In reply to Sasha Levin

From: Jeffrey Hugo <[email protected]>

[ Upstream commit 78e31c42261779a01bc73472d0f65f15378e9de3 ]

On msm8998, vblank timeouts are observed because the DSI controller is not reset properly, which ends up stalling the MDP. This is because the reset logic is not correct per the hardware documentation.

The documentation states that after asserting reset, software should wait some time (no indication of how long), or poll the status register until it returns 0 before deasserting reset.

wmb() is insufficient for this purpose since it just ensures ordering, not timing between writes. Since asserting and deasserting reset occurs on the same register, ordering is already guaranteed by the architecture, making the wmb extraneous.

Since we would define a timeout for polling the status register to avoid a possible infinite loop, lets just use a static delay of 20 ms, since 16.666 ms is the time available to process one frame at 60 fps.

Fixes a689554ba6ed ("drm/msm: Initial add DSI connector support")
Cc Hai Li <[email protected]>
Cc Rob Clark <[email protected]>
[seanpaul renamed RESET_DELAY to DSI_RESET_TOGGLE_DELAY_MS]
---
 drivers/gpu/drm/msm/dsi/dsi_host.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/gpu/drm/msm/dsi/dsi_host.c +4 -2
index cc4ea5502d6c..3b78bca0bb4d 100644
@@ -34,6 +34,8 @@
 #include "dsi_cfg.h"
 #include "msm_kms.h"
 
+#define DSI_RESET_TOGGLE_DELAY_MS 20
+
 static int dsi_get_version(const void __iomem *base, u32 *major, u32 *minor)
 {
 	u32 ver;
@@ -994,7 +996,7 @@ static void dsi_sw_reset(struct msm_dsi_host *msm_host)
 	wmb(); /* clocks need to be enabled before reset */
 
 	dsi_write(msm_host, REG_DSI_RESET, 1);
-	wmb(); /* make sure reset happen */
+	msleep(DSI_RESET_TOGGLE_DELAY_MS); /* make sure reset happen */
 	dsi_write(msm_host, REG_DSI_RESET, 0);
 }
 
@@ -1402,7 +1404,7 @@ static void dsi_sw_reset_restore(struct msm_dsi_host *msm_host)
 
 	/* dsi controller can only be reset while clocks are running */
 	dsi_write(msm_host, REG_DSI_RESET, 1);
-	wmb();	/* make sure reset happen */
+	msleep(DSI_RESET_TOGGLE_DELAY_MS); /* make sure reset happen */
 	dsi_write(msm_host, REG_DSI_RESET, 0);
 	wmb();	/* controller out of reset */
 	dsi_write(msm_host, REG_DSI_CTRL, data0);
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel
author maintainer
#340 lore raw

PATCH 596/671 vhost/test: stop device before reset

In reply to Sasha Levin

From: "Michael S. Tsirkin" <[email protected]>

[ Upstream commit 245cdd9fbd396483d501db83047116e2530f245f ]

When device stop was moved out of reset, test device wasn't updated to stop before reset, this resulted in a use after free. Fix by invoking stop appropriately.

Fixes b211616d7125 ("vhost: move -net specific code out")
---
 drivers/vhost/test.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/vhost/test.c +2 -0
index a9be2d8e98df..55090d9f9de0 100644
@@ -162,6 +162,7 @@ static int vhost_test_release(struct inode *inode, struct file *f)
 
 	vhost_test_stop(n, &private);
 	vhost_test_flush(n);
+	vhost_dev_stop(&n->dev);
 	vhost_dev_cleanup(&n->dev);
 	/* We do an extra flush before freeing memory,
 	 * since jobs can re-queue themselves. */
@@ -238,6 +239,7 @@ static long vhost_test_reset_owner(struct vhost_test *n)
 	}
 	vhost_test_stop(n, &priv);
 	vhost_test_flush(n);
+	vhost_dev_stop(&n->dev);
 	vhost_dev_reset_owner(&n->dev, umem);
 done:
 	mutex_unlock(&n->dev.mutex);
Signature · 1 line
2.20.1
author maintainer
#341 lore raw

PATCH 597/671 dmaengine: imx-sdma: fix size check for sdma script_number

In reply to Sasha Levin

From: Robin Gong <[email protected]>

[ Upstream commit bd73dfabdda280fc5f05bdec79b6721b4b2f035f ]

Illegal memory will be touch if SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V3
(41) exceed the size of structure sdma_script_start_addrs(40), thus cause memory corrupt such as slob block header so that kernel trap into while() loop forever in slob_free(). Please refer to below code piece in imx-sdma.c:
for (i = 0; i < sdma->script_number; i++)

	if (addr_arr[i] > 0)
		saddr_arr[i] = addr_arr[i]; /* memory corrupt here */

That issue was brought by commit a572460be9cf ("dmaengine: imx-sdma: Add support for version 3 firmware") because SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V3 (38->41 3 scripts added) not align with script number added in sdma_script_start_addrs(2 scripts).

Fixes a572460be9cf ("dmaengine: imx-sdma: Add support for version 3 firmware")
[vkoul: update the patch title]
---
 drivers/dma/imx-sdma.c                     | 8 ++++++++
 include/linux/platform_data/dma-imx-sdma.h | 3 +++
 2 files changed, 11 insertions(+)
drivers/dma/imx-sdma.c +8 -0
index 3f5a01cb4ab4..ceb82e74f5b4 100644
@@ -1662,6 +1662,14 @@ static void sdma_add_scripts(struct sdma_engine *sdma,
 	if (!sdma->script_number)
 		sdma->script_number = SDMA_SCRIPT_ADDRS_ARRAY_SIZE_V1;
 
+	if (sdma->script_number > sizeof(struct sdma_script_start_addrs)
+				  / sizeof(s32)) {
+		dev_err(sdma->dev,
+			"SDMA script number %d not match with firmware.\n",
+			sdma->script_number);
+		return;
+	}
+
 	for (i = 0; i < sdma->script_number; i++)
 		if (addr_arr[i] > 0)
 			saddr_arr[i] = addr_arr[i];
include/linux/platform_data/dma-imx-sdma.h +3 -0
index 6eaa53cef0bd..30e676b36b24 100644
@@ -51,7 +51,10 @@ struct sdma_script_start_addrs {
 	/* End of v2 array */
 	s32 zcanfd_2_mcu_addr;
 	s32 zqspi_2_mcu_addr;
+	s32 mcu_2_ecspi_addr;
 	/* End of v3 array */
+	s32 mcu_2_zqspi_addr;
+	/* End of v4 array */
 };
 
 /**
Signature · 1 line
2.20.1
author maintainer
#342 lore raw

PATCH 598/671 firmware: dmi: Fix unlikely out-of-bounds read in save_mem_devices

In reply to Sasha Levin

From: Jean Delvare <[email protected]>

[ Upstream commit 81dde26de9c08bb04c4962a15608778aaffb3cf9 ]

Before reading the Extended Size field, we should ensure it fits in the DMI record. There is already a record length check but it does not cover that field.

It would take a seriously corrupted DMI table to hit that bug, so no need to worry, but we should still fix it.

Fixes 6deae96b42eb ("firmware, DMI: Add function to look up a handle and return DIMM size")
Cc Tony Luck <[email protected]>
Cc Borislav Petkov <[email protected]>
---
 drivers/firmware/dmi_scan.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/firmware/dmi_scan.c +1 -1
index f2483548cde9..0dc0c78f1fdb 100644
@@ -407,7 +407,7 @@ static void __init save_mem_devices(const struct dmi_header *dm, void *v)
 		bytes = ~0ull;
 	else if (size & 0x8000)
 		bytes = (u64)(size & 0x7fff) << 10;
-	else if (size != 0x7fff)
+	else if (size != 0x7fff || dm->length < 0x20)
 		bytes = (u64)size << 20;
 	else
 		bytes = (u64)get_unaligned((u32 *)&d[0x1C]) << 20;
Signature · 1 line
2.20.1
author maintainer
#343 lore raw

PATCH 599/671 arm64: hibernate: check pgd table allocation

In reply to Sasha Levin

From: Pavel Tatashin <[email protected]>

[ Upstream commit 8c551f919a73c1dfa690a70a691be1da394145e8 ]

There is a bug in create_safe_exec_page(), when page table is allocated it is not checked that table is allocated successfully:

But it is dereferenced in: pgd_none(READ_ONCE(*pgdp)). Check that allocation was successful.

Fixes 82869ac57b5d ("arm64: kernel: Add support for hibernate/suspend-to-disk")
---
 arch/arm64/kernel/hibernate.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
arch/arm64/kernel/hibernate.c +8 -1
index 9859e1178e6b..dbeeeffdb9c9 100644
@@ -202,6 +202,7 @@ static int create_safe_exec_page(void *src_start, size_t length,
 				 gfp_t mask)
 {
 	int rc = 0;
+	pgd_t *trans_pgd;
 	pgd_t *pgdp;
 	pud_t *pudp;
 	pmd_t *pmdp;
@@ -216,7 +217,13 @@ static int create_safe_exec_page(void *src_start, size_t length,
 	memcpy((void *)dst, src_start, length);
 	__flush_icache_range(dst, dst + length);
 
-	pgdp = pgd_offset_raw(allocator(mask), dst_addr);
+	trans_pgd = allocator(mask);
+	if (!trans_pgd) {
+		rc = -ENOMEM;
+		goto out;
+	}
+
+	pgdp = pgd_offset_raw(trans_pgd, dst_addr);
 	if (pgd_none(READ_ONCE(*pgdp))) {
 		pudp = allocator(mask);
 		if (!pudp) {
Signature · 1 line
2.20.1
author maintainer
#344 lore raw

PATCH 600/671 net: netem: fix error path for corrupted GSO frames

In reply to Sasha Levin

From: Jakub Kicinski <[email protected]>

[ Upstream commit a7fa12d15855904aff1716e1fc723c03ba38c5cc ]

To corrupt a GSO frame we first perform segmentation. We then proceed using the first segment instead of the full GSO skb and requeue the rest of the segments as separate packets.

If there are any issues with processing the first segment we still want to process the rest, therefore we jump to the finish_segs label.

Commit 177b8007463c ("net: netem: fix backlog accounting for corrupted GSO frames") started using the pointer to the first segment in the "rest of segments processing", but as mentioned above the first segment may had already been freed at this point.

Backlog corrections for parent qdiscs have to be adjusted.

Fixes 177b8007463c ("net: netem: fix backlog accounting for corrupted GSO frames")
---
 net/sched/sch_netem.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
net/sched/sch_netem.c +6 -3
index 1cd7266140e6..7660aa5b80da 100644
@@ -509,6 +509,7 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch,
 		if (skb->ip_summed == CHECKSUM_PARTIAL &&
 		    skb_checksum_help(skb)) {
 			qdisc_drop(skb, sch, to_free);
+			skb = NULL;
 			goto finish_segs;
 		}
 
@@ -584,9 +585,10 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch,
 finish_segs:
 	if (segs) {
 		unsigned int len, last_len;
-		int nb = 0;
+		int nb;
 
-		len = skb->len;
+		len = skb ? skb->len : 0;
+		nb = skb ? 1 : 0;
 
 		while (segs) {
 			skb2 = segs->next;
@@ -603,7 +605,8 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch,
 			}
 			segs = skb2;
 		}
-		qdisc_tree_reduce_backlog(sch, -nb, prev_len - len);
+		/* Parent qdiscs accounted for 1 skb of size @prev_len */
+		qdisc_tree_reduce_backlog(sch, -(nb - 1), -(len - prev_len));
 	}
 	return NET_XMIT_SUCCESS;
 }
Signature · 1 line
2.20.1
author maintainer
#345 lore raw

PATCH 601/671 net: netem: correct the parent's backlog when corrupted packet was dropped

In reply to Sasha Levin

From: Jakub Kicinski <[email protected]>

[ Upstream commit e0ad032e144731a5928f2d75e91c2064ba1a764c ]

If packet corruption failed we jump to finish_segs and return NET_XMIT_SUCCESS. Seeing success will make the parent qdisc increment its backlog, that's incorrect - we need to return NET_XMIT_DROP.

Fixes 6071bd1aa13e ("netem: Segment GSO packets on enqueue")
---
 net/sched/sch_netem.c | 2 ++
 1 file changed, 2 insertions(+)
net/sched/sch_netem.c +2 -0
index 7660aa5b80da..014a28d8dd4f 100644
@@ -607,6 +607,8 @@ static int netem_enqueue(struct sk_buff *skb, struct Qdisc *sch,
 		}
 		/* Parent qdiscs accounted for 1 skb of size @prev_len */
 		qdisc_tree_reduce_backlog(sch, -(nb - 1), -(len - prev_len));
+	} else if (!skb) {
+		return NET_XMIT_DROP;
 	}
 	return NET_XMIT_SUCCESS;
 }
Signature · 1 line
2.20.1
author maintainer
#346 lore raw

PATCH 602/671 xsk: Fix registration of Rx-only sockets

In reply to Sasha Levin

From: Magnus Karlsson <[email protected]>

[ Upstream commit 2afd23f78f39da84937006ecd24aa664a4ab052b ]

Having Rx-only AF_XDP sockets can potentially lead to a crash in the system by a NULL pointer dereference in xsk_umem_consume_tx(). This function iterates through a list of all sockets tied to a umem and checks if there are any packets to send on the Tx ring. Rx-only sockets do not have a Tx ring, so this will cause a NULL pointer dereference. This will happen if you have registered one or more Rx-only sockets to a umem and the driver is checking the Tx ring even on Rx, or if the XDP_SHARED_UMEM mode is used and there is a mix of Rx-only and other sockets tied to the same umem.

Fixed by only putting sockets with a Tx component on the list that xsk_umem_consume_tx() iterates over.

---
 net/xdp/xdp_umem.c | 6 ++++++
 1 file changed, 6 insertions(+)
net/xdp/xdp_umem.c +6 -0
index d9117ab035f7..556a649512b6 100644
@@ -23,6 +23,9 @@ void xdp_add_sk_umem(struct xdp_umem *umem, struct xdp_sock *xs)
 {
 	unsigned long flags;
 
+	if (!xs->tx)
+		return;
+
 	spin_lock_irqsave(&umem->xsk_list_lock, flags);
 	list_add_rcu(&xs->list, &umem->xsk_list);
 	spin_unlock_irqrestore(&umem->xsk_list_lock, flags);
@@ -32,6 +35,9 @@ void xdp_del_sk_umem(struct xdp_umem *umem, struct xdp_sock *xs)
 {
 	unsigned long flags;
 
+	if (!xs->tx)
+		return;
+
 	spin_lock_irqsave(&umem->xsk_list_lock, flags);
 	list_del_rcu(&xs->list);
 	spin_unlock_irqrestore(&umem->xsk_list_lock, flags);
Signature · 1 line
2.20.1