Skip to contentSkip to conversation

ARM: riscpc: fix DMA

PATCH New Sasha Levin 1 participant 47 messages Jul 26, 2019

Subsystems (17)
lore
Share & tools
lore

Sasha Levin

author maintainer
#0 lore raw

PATCH 1/47 ARM: riscpc: fix DMA

From: Russell King <[email protected]>

[ Upstream commit ffd9a1ba9fdb7f2bd1d1ad9b9243d34e96756ba2 ]

DMA got broken a while back in two different ways:
1) a change in the behaviour of disable_irq() to wait for the interrupt to finish executing causes us to deadlock at the end of DMA.
2) a change to avoid modifying the scatterlist left the first transfer uninitialised.

DMA is only used with expansion cards, so has gone unnoticed.

Fixes fa4e99899932 ("[ARM] dma: RiscPC: don't modify DMA SG entries")
---
 arch/arm/mach-rpc/dma.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
arch/arm/mach-rpc/dma.c +4 -1
index fb48f3141fb4..c4c96661eb89 100644
@@ -131,7 +131,7 @@ static irqreturn_t iomd_dma_handle(int irq, void *dev_id)
 	} while (1);
 
 	idma->state = ~DMA_ST_AB;
-	disable_irq(irq);
+	disable_irq_nosync(irq);
 
 	return IRQ_HANDLED;
 }
@@ -174,6 +174,9 @@ static void iomd_enable_dma(unsigned int chan, dma_t *dma)
 				DMA_FROM_DEVICE : DMA_TO_DEVICE);
 		}
 
+		idma->dma_addr = idma->dma.sg->dma_address;
+		idma->dma_len = idma->dma.sg->length;
+
 		iomd_writeb(DMA_CR_C, dma_base + CR);
 		idma->state = DMA_ST_AB;
 	}
Signature · 1 line
2.20.1
author maintainer
#1 lore raw

PATCH 2/47 ARM: dts: rockchip: Make rk3288-veyron-minnie run at hs200

From: Douglas Anderson <[email protected]>

[ Upstream commit 1c0479023412ab7834f2e98b796eb0d8c627cd62 ]

As some point hs200 was failing on rk3288-veyron-minnie. See commit 984926781122 ("ARM: dts: rockchip: temporarily remove emmc hs200 speed from rk3288 minnie"). Although I didn't track down exactly when it started working, it seems to work OK now, so let's turn it back on.

To test this, I booted from SD card and then used this script to stress the enumeration process after fixing a memory leak [1]:
cd /sys/bus/platform/drivers/dwmmc_rockchip
for i in $(seq 1 3000); do
echo "========================" $i
echo ff0f0000.dwmmc > unbind
sleep .5
echo ff0f0000.dwmmc > bind
while true; do
if [ -e /dev/mmcblk2 ]; then
break;
fi
sleep .1
done
done

It worked fine.

[1] https://lkml.kernel.org/r/20190503233526.226272-1-dianders-F7+t8E8rja9g9hUCZPvPmw@public.gmane.org

---
 arch/arm/boot/dts/rk3288-veyron-minnie.dts | 4 ----
 1 file changed, 4 deletions(-)
arch/arm/boot/dts/rk3288-veyron-minnie.dts +0 -4
index f95d0c5fcf71..6e8946052c78 100644
@@ -90,10 +90,6 @@
 	pwm-off-delay-ms = <200>;
 };
 
-&emmc {
-	/delete-property/mmc-hs200-1_8v;
-};
-
 &gpio_keys {
 	pinctrl-0 = <&pwr_key_l &ap_lid_int_l &volum_down_l &volum_up_l>;
 
Signature · 1 line
2.20.1
author maintainer
#2 lore raw

PATCH 3/47 ARM: dts: rockchip: Make rk3288-veyron-mickey's emmc work again

In reply to Sasha Levin

From: Douglas Anderson <[email protected]>

[ Upstream commit 99fa066710f75f18f4d9a5bc5f6a711968a581d5 ]

When I try to boot rk3288-veyron-mickey I totally fail to make the eMMC work. Specifically my logs (on Chrome OS 4.19):

mmc_host mmc1: card is non-removable.
mmc_host mmc1: Bus speed (slot 0) = 400000Hz (slot req 400000Hz, actual 400000HZ div = 0) mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 52000000Hz, actual 50000000HZ div = 0)
mmc1: switch to bus width 8 failed
mmc1: switch to bus width 4 failed
mmc1: new high speed MMC card at address 0001
mmcblk1: mmc1:0001 HAG2e 14.7 GiB
mmcblk1boot0: mmc1:0001 HAG2e partition 1 4.00 MiB
mmcblk1boot1: mmc1:0001 HAG2e partition 2 4.00 MiB
mmcblk1rpmb: mmc1:0001 HAG2e partition 3 4.00 MiB, chardev (243:0)
mmc_host mmc1: Bus speed (slot 0) = 400000Hz (slot req 400000Hz, actual 400000HZ div = 0) mmc_host mmc1: Bus speed (slot 0) = 50000000Hz (slot req 52000000Hz, actual 50000000HZ div = 0)
mmc1: switch to bus width 8 failed
mmc1: switch to bus width 4 failed
mmc1: tried to HW reset card, got error -110
mmcblk1: error -110 requesting status
mmcblk1: recovery failed!
print_req_error: I/O error, dev mmcblk1, sector 0
...

When I remove the '/delete-property/mmc-hs200-1_8v' then everything is hunky dory.

That line comes from the original submission of the mickey dts upstream, so presumably at the time the HS200 was failing and just enumerating things as a high speed device was fine. ...or maybe it's just that some mickey devices work when enumerating at "high speed", just not mine?

In any case, hs200 seems good now. Let's turn it on.

---
 arch/arm/boot/dts/rk3288-veyron-mickey.dts | 4 ----
 1 file changed, 4 deletions(-)
arch/arm/boot/dts/rk3288-veyron-mickey.dts +0 -4
index 1e0158acf895..a593d0a998fc 100644
@@ -124,10 +124,6 @@
 	};
 };
 
-&emmc {
-	/delete-property/mmc-hs200-1_8v;
-};
-
 &i2c2 {
 	status = "disabled";
 };
Signature · 1 line
2.20.1
author maintainer
#3 lore raw

PATCH 4/47 ARM: dts: rockchip: Mark that the rk3288 timer might stop in suspend

In reply to Sasha Levin

From: Douglas Anderson <[email protected]>

[ Upstream commit 8ef1ba39a9fa53d2205e633bc9b21840a275908e ]

This is similar to commit e6186820a745 ("arm64: dts: rockchip: Arch counter doesn't tick in system suspend"). Specifically on the rk3288 it can be seen that the timer stops ticking in suspend if we end up running through the "osc_disable" path in rk3288_slp_mode_set(). In that path the 24 MHz clock will turn off and the timer stops.

To test this, I ran this on a Chrome OS filesystem:
before=$(date); \
suspend_stress_test -c1 --suspend_min=30 --suspend_max=31; \
echo ${before}; date

...and I found that unless I plug in a device that requests USB wakeup to be active that the two calls to "date" would show that fewer than 30 seconds passed.

NOTE: deep suspend (where the 24 MHz clock gets disabled) isn't supported yet on upstream Linux so this was tested on a downstream kernel.

---
 arch/arm/boot/dts/rk3288.dtsi | 1 +
 1 file changed, 1 insertion(+)
arch/arm/boot/dts/rk3288.dtsi +1 -0
index c706adf4aed2..440d6783faca 100644
@@ -227,6 +227,7 @@
 			     <GIC_PPI 11 (GIC_CPU_MASK_SIMPLE(4) | IRQ_TYPE_LEVEL_HIGH)>,
 			     <GIC_PPI 10 (GIC_CPU_MASK_SIMPLE(4) | IRQ_TYPE_LEVEL_HIGH)>;
 		clock-frequency = <24000000>;
+		arm,no-tick-in-suspend;
 	};
 
 	timer: timer@ff810000 {
Signature · 1 line
2.20.1
author maintainer
#4 lore raw

PATCH 5/47 ftrace: Enable trampoline when rec count returns back to one

In reply to Sasha Levin

From: Cheng Jian <[email protected]>

[ Upstream commit a124692b698b00026a58d89831ceda2331b2e1d0 ]

Custom trampolines can only be enabled if there is only a single ops attached to it. If there's only a single callback registered to a function, and the ops has a trampoline registered for it, then we can call the trampoline directly. This is very useful for improving the performance of ftrace and livepatch.

If more than one callback is registered to a function, the general trampoline is used, and the custom trampoline is not restored back to the direct call even if all the other callbacks were unregistered and we are back to one callback for the function.

To fix this, set FTRACE_FL_TRAMP flag if rec count is decremented to one, and the ops that left has a trampoline.

Testing After this patch :

insmod livepatch_unshare_files.ko
cat /sys/kernel/debug/tracing/enabled_functions

	unshare_files (1) R I	tramp: 0xffffffffc0000000(klp_ftrace_handler+0x0/0xa0) ->ftrace_ops_assist_func+0x0/0xf0

echo unshare_files > /sys/kernel/debug/tracing/set_ftrace_filter echo function > /sys/kernel/debug/tracing/current_tracer cat /sys/kernel/debug/tracing/enabled_functions

	unshare_files (2) R I ->ftrace_ops_list_func+0x0/0x150

echo nop > /sys/kernel/debug/tracing/current_tracer cat /sys/kernel/debug/tracing/enabled_functions

	unshare_files (1) R I	tramp: 0xffffffffc0000000(klp_ftrace_handler+0x0/0xa0) ->ftrace_ops_assist_func+0x0/0xf0
---
 kernel/trace/ftrace.c | 28 +++++++++++++++-------------
 1 file changed, 15 insertions(+), 13 deletions(-)
kernel/trace/ftrace.c +15 -13
index 118ecce14386..d9dd709b3c12 100644
@@ -1647,6 +1647,11 @@ static bool test_rec_ops_needs_regs(struct dyn_ftrace *rec)
 	return  keep_regs;
 }
 
+static struct ftrace_ops *
+ftrace_find_tramp_ops_any(struct dyn_ftrace *rec);
+static struct ftrace_ops *
+ftrace_find_tramp_ops_next(struct dyn_ftrace *rec, struct ftrace_ops *ops);
+
 static bool __ftrace_hash_rec_update(struct ftrace_ops *ops,
 				     int filter_hash,
 				     bool inc)
@@ -1775,15 +1780,17 @@ static bool __ftrace_hash_rec_update(struct ftrace_ops *ops,
 			}
 
 			/*
-			 * If the rec had TRAMP enabled, then it needs to
-			 * be cleared. As TRAMP can only be enabled iff
-			 * there is only a single ops attached to it.
-			 * In otherwords, always disable it on decrementing.
-			 * In the future, we may set it if rec count is
-			 * decremented to one, and the ops that is left
-			 * has a trampoline.
+			 * The TRAMP needs to be set only if rec count
+			 * is decremented to one, and the ops that is
+			 * left has a trampoline. As TRAMP can only be
+			 * enabled if there is only a single ops attached
+			 * to it.
 			 */
-			rec->flags &= ~FTRACE_FL_TRAMP;
+			if (ftrace_rec_count(rec) == 1 &&
+			    ftrace_find_tramp_ops_any(rec))
+				rec->flags |= FTRACE_FL_TRAMP;
+			else
+				rec->flags &= ~FTRACE_FL_TRAMP;
 
 			/*
 			 * flags will be cleared in ftrace_check_record()
@@ -1976,11 +1983,6 @@ static void print_ip_ins(const char *fmt, const unsigned char *p)
 		printk(KERN_CONT "%s%02x", i ? ":" : "", p[i]);
 }
 
-static struct ftrace_ops *
-ftrace_find_tramp_ops_any(struct dyn_ftrace *rec);
-static struct ftrace_ops *
-ftrace_find_tramp_ops_next(struct dyn_ftrace *rec, struct ftrace_ops *ops);
-
 enum ftrace_bug_type ftrace_bug_type;
 const void *ftrace_expected;
 
Signature · 1 line
2.20.1
author maintainer
#5 lore raw

PATCH 6/47 dmaengine: tegra-apb: Error out if DMA_PREP_INTERRUPT flag is unset

In reply to Sasha Levin

From: Dmitry Osipenko <[email protected]>

[ Upstream commit dc161064beb83c668e0f85766b92b1e7ed186e58 ]

Apparently driver was never tested with DMA_PREP_INTERRUPT flag being unset since it completely disables interrupt handling instead of skipping the callbacks invocations, hence putting channel into unusable state.

The flag is always set by all of kernel drivers that use APB DMA, so let's error out in otherwise case for consistency. It won't be difficult to support that case properly if ever will be needed.

---
 drivers/dma/tegra20-apb-dma.c | 12 ++++++++++--
 1 file changed, 10 insertions(+), 2 deletions(-)
drivers/dma/tegra20-apb-dma.c +10 -2
index 8219ab88a507..fb23993430d3 100644
@@ -981,8 +981,12 @@ static struct dma_async_tx_descriptor *tegra_dma_prep_slave_sg(
 		csr |= tdc->slave_id << TEGRA_APBDMA_CSR_REQ_SEL_SHIFT;
 	}
 
-	if (flags & DMA_PREP_INTERRUPT)
+	if (flags & DMA_PREP_INTERRUPT) {
 		csr |= TEGRA_APBDMA_CSR_IE_EOC;
+	} else {
+		WARN_ON_ONCE(1);
+		return NULL;
+	}
 
 	apb_seq |= TEGRA_APBDMA_APBSEQ_WRAP_WORD_1;
 
@@ -1124,8 +1128,12 @@ static struct dma_async_tx_descriptor *tegra_dma_prep_dma_cyclic(
 		csr |= tdc->slave_id << TEGRA_APBDMA_CSR_REQ_SEL_SHIFT;
 	}
 
-	if (flags & DMA_PREP_INTERRUPT)
+	if (flags & DMA_PREP_INTERRUPT) {
 		csr |= TEGRA_APBDMA_CSR_IE_EOC;
+	} else {
+		WARN_ON_ONCE(1);
+		return NULL;
+	}
 
 	apb_seq |= TEGRA_APBDMA_APBSEQ_WRAP_WORD_1;
 
Signature · 1 line
2.20.1
author maintainer
#6 lore raw

PATCH 7/47 arm64: dts: rockchip: fix isp iommu clocks and power domain

In reply to Sasha Levin

From: Helen Koike <[email protected]>

[ Upstream commit c432a29d3fc9ee928caeca2f5cf68b3aebfa6817 ]

isp iommu requires wrapper variants of the clocks.
noc variants are always on and using the wrapper variants will activate {A,H}CLK_ISP{0,1} due to the hierarchy.

Tested using the pending isp patch set (which is not upstream yet). Without this patch, streaming from the isp stalls.

Also add the respective power domain and remove the "disabled" status.

Refer:
RK3399 TRM v1.4 Fig. 2-4 RK3399 Clock Architecture Diagram RK3399 TRM v1.4 Fig. 8-1 RK3399 Power Domain Partition

---
 arch/arm64/boot/dts/rockchip/rk3399.dtsi | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
arch/arm64/boot/dts/rockchip/rk3399.dtsi +4 -4
index df7e62d9a670..cea44a7c7cf9 100644
@@ -1643,11 +1643,11 @@
 		reg = <0x0 0xff914000 0x0 0x100>, <0x0 0xff915000 0x0 0x100>;
 		interrupts = <GIC_SPI 43 IRQ_TYPE_LEVEL_HIGH 0>;
 		interrupt-names = "isp0_mmu";
-		clocks = <&cru ACLK_ISP0_NOC>, <&cru HCLK_ISP0_NOC>;
+		clocks = <&cru ACLK_ISP0_WRAPPER>, <&cru HCLK_ISP0_WRAPPER>;
 		clock-names = "aclk", "iface";
 		#iommu-cells = <0>;
+		power-domains = <&power RK3399_PD_ISP0>;
 		rockchip,disable-mmu-reset;
-		status = "disabled";
 	};
 
 	isp1_mmu: iommu@ff924000 {
@@ -1655,11 +1655,11 @@
 		reg = <0x0 0xff924000 0x0 0x100>, <0x0 0xff925000 0x0 0x100>;
 		interrupts = <GIC_SPI 44 IRQ_TYPE_LEVEL_HIGH 0>;
 		interrupt-names = "isp1_mmu";
-		clocks = <&cru ACLK_ISP1_NOC>, <&cru HCLK_ISP1_NOC>;
+		clocks = <&cru ACLK_ISP1_WRAPPER>, <&cru HCLK_ISP1_WRAPPER>;
 		clock-names = "aclk", "iface";
 		#iommu-cells = <0>;
+		power-domains = <&power RK3399_PD_ISP1>;
 		rockchip,disable-mmu-reset;
-		status = "disabled";
 	};
 
 	hdmi_sound: hdmi-sound {
Signature · 1 line
2.20.1
author maintainer
#7 lore raw

PATCH 8/47 kernel/module.c: Only return -EEXIST for modules that have finished loading

In reply to Sasha Levin

From: Prarit Bhargava <[email protected]>

[ Upstream commit 6e6de3dee51a439f76eb73c22ae2ffd2c9384712 ]

Microsoft HyperV disables the X86_FEATURE_SMCA bit on AMD systems, and linux guests boot with repeated errors:

amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_report_gart_errors (err -2)
amd64_edac_mod: Unknown symbol amd_unregister_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_register_ecc_decoder (err -2)
amd64_edac_mod: Unknown symbol amd_report_gart_errors (err -2)

The warnings occur because the module code erroneously returns -EEXIST for modules that have failed to load and are in the process of being removed from the module list.

module amd64_edac_mod has a dependency on module edac_mce_amd. Using modules.dep, systemd will load edac_mce_amd for every request of amd64_edac_mod. When the edac_mce_amd module loads, the module has state MODULE_STATE_UNFORMED and once the module load fails and the state becomes MODULE_STATE_GOING. Another request for edac_mce_amd module executes and add_unformed_module() will erroneously return -EEXIST even though the previous instance of edac_mce_amd has MODULE_STATE_GOING. Upon receiving -EEXIST, systemd attempts to load amd64_edac_mod, which fails because of unknown symbols from edac_mce_amd.

add_unformed_module() must wait to return for any case other than MODULE_STATE_LIVE to prevent a race between multiple loads of dependent modules.

---
 kernel/module.c | 6 ++----
 1 file changed, 2 insertions(+), 4 deletions(-)
kernel/module.c +2 -4
index b8f37376856b..3fda10c549a2 100644
@@ -3388,8 +3388,7 @@ static bool finished_loading(const char *name)
 	sched_annotate_sleep();
 	mutex_lock(&module_mutex);
 	mod = find_module_all(name, strlen(name), true);
-	ret = !mod || mod->state == MODULE_STATE_LIVE
-		|| mod->state == MODULE_STATE_GOING;
+	ret = !mod || mod->state == MODULE_STATE_LIVE;
 	mutex_unlock(&module_mutex);
 
 	return ret;
@@ -3559,8 +3558,7 @@ static int add_unformed_module(struct module *mod)
 	mutex_lock(&module_mutex);
 	old = find_module_all(mod->name, strlen(mod->name), true);
 	if (old != NULL) {
-		if (old->state == MODULE_STATE_COMING
-		    || old->state == MODULE_STATE_UNFORMED) {
+		if (old->state != MODULE_STATE_LIVE) {
 			/* Wait in case it fails to load. */
 			mutex_unlock(&module_mutex);
 			err = wait_event_interruptible(module_wq,
Signature · 1 line
2.20.1
author maintainer
#8 lore raw

PATCH 9/47 firmware/psci: psci_checker: Park kthreads before stopping them

In reply to Sasha Levin

From: Jean-Philippe Brucker <[email protected]>

[ Upstream commit 92e074acf6f7694e96204265eb18ac113f546e80 ]

Since commit 85f1abe0019f ("kthread, sched/wait: Fix kthread_parkme() completion issue"), kthreads that are bound to a CPU must be parked before being stopped. At the moment the PSCI checker calls kthread_stop() directly on the suspend kthread, which triggers the following warning:

[    6.068288] WARNING: CPU: 1 PID: 1 at kernel/kthread.c:398 __kthread_bind_mask+0x20/0x78
               ...
[    6.190151] Call trace:
[    6.192566]  __kthread_bind_mask+0x20/0x78
[    6.196615]  kthread_unpark+0x74/0x80
[    6.200235]  kthread_stop+0x44/0x1d8
[    6.203769]  psci_checker+0x3bc/0x484
[    6.207389]  do_one_initcall+0x48/0x260
[    6.211180]  kernel_init_freeable+0x2c8/0x368
[    6.215488]  kernel_init+0x10/0x100
[    6.218935]  ret_from_fork+0x10/0x1c
[    6.222467] ---[ end trace e05e22863d043cd3 ]---

kthread_unpark() tries to bind the thread to its CPU and aborts with a WARN() if the thread wasn't in TASK_PARKED state. Park the kthreads before stopping them.

Fixes 85f1abe0019f ("kthread, sched/wait: Fix kthread_parkme() completion issue")
---
 drivers/firmware/psci_checker.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/firmware/psci_checker.c +6 -4
index 346943657962..cbd53cb1b2d4 100644
@@ -366,16 +366,16 @@ static int suspend_test_thread(void *arg)
 	for (;;) {
 		/* Needs to be set first to avoid missing a wakeup. */
 		set_current_state(TASK_INTERRUPTIBLE);
-		if (kthread_should_stop()) {
-			__set_current_state(TASK_RUNNING);
+		if (kthread_should_park())
 			break;
-		}
 		schedule();
 	}
 
 	pr_info("CPU %d suspend test results: success %d, shallow states %d, errors %d\n",
 		cpu, nb_suspend, nb_shallow_sleep, nb_err);
 
+	kthread_parkme();
+
 	return nb_err;
 }
 
@@ -440,8 +440,10 @@ static int suspend_tests(void)
 
 
 	/* Stop and destroy all threads, get return status. */
-	for (i = 0; i < nb_threads; ++i)
+	for (i = 0; i < nb_threads; ++i) {
+		err += kthread_park(threads[i]);
 		err += kthread_stop(threads[i]);
+	}
  out:
 	cpuidle_resume_and_unlock();
 	kfree(threads);
Signature · 1 line
2.20.1
author maintainer
#9 lore raw

PATCH 10/47 MIPS: lantiq: Fix bitfield masking

In reply to Sasha Levin

From: Petr Cvek <[email protected]>

[ Upstream commit ba1bc0fcdeaf3bf583c1517bd2e3e29cf223c969 ]

The modification of EXIN register doesn't clean the bitfield before the writing of a new value. After a few modifications the bitfield would accumulate only '1's.

---
 arch/mips/lantiq/irq.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)
arch/mips/lantiq/irq.c +3 -2
index c4ef1c31e0c4..37caeadb2964 100644
@@ -156,8 +156,9 @@ static int ltq_eiu_settype(struct irq_data *d, unsigned int type)
 			if (edge)
 				irq_set_handler(d->hwirq, handle_edge_irq);
 
-			ltq_eiu_w32(ltq_eiu_r32(LTQ_EIU_EXIN_C) |
-				(val << (i * 4)), LTQ_EIU_EXIN_C);
+			ltq_eiu_w32((ltq_eiu_r32(LTQ_EIU_EXIN_C) &
+				    (~(7 << (i * 4)))) | (val << (i * 4)),
+				    LTQ_EIU_EXIN_C);
 		}
 	}
 
Signature · 1 line
2.20.1
author maintainer
#10 lore raw

PATCH 11/47 dmaengine: rcar-dmac: Reject zero-length slave DMA requests

In reply to Sasha Levin

From: Geert Uytterhoeven <[email protected]>

[ Upstream commit 78efb76ab4dfb8f74f290ae743f34162cd627f19 ]

While the .device_prep_slave_sg() callback rejects empty scatterlists, it still accepts single-entry scatterlists with a zero-length segment. These may happen if a driver calls dmaengine_prep_slave_single() with a zero len parameter. The corresponding DMA request will never complete, leading to messages like:

rcar-dmac e7300000.dma-controller: Channel Address Error happen

and DMA timeouts.

Although requesting a zero-length DMA request is a driver bug, rejecting it early eases debugging. Note that the .device_prep_dma_memcpy() callback already rejects requests to copy zero bytes.

---
 drivers/dma/sh/rcar-dmac.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/dma/sh/rcar-dmac.c +1 -1
index 0b05a1e08d21..041ce864097e 100644
@@ -1164,7 +1164,7 @@ rcar_dmac_prep_slave_sg(struct dma_chan *chan, struct scatterlist *sgl,
 	struct rcar_dmac_chan *rchan = to_rcar_dmac_chan(chan);
 
 	/* Someone calling slave DMA on a generic channel? */
-	if (rchan->mid_rid < 0 || !sg_len) {
+	if (rchan->mid_rid < 0 || !sg_len || !sg_dma_len(sgl)) {
 		dev_warn(chan->device->dev,
 			 "%s: bad parameter: len=%d, id=%d\n",
 			 __func__, sg_len, rchan->mid_rid);
Signature · 1 line
2.20.1
author maintainer
#11 lore raw

PATCH 12/47 clk: tegra210: fix PLLU and PLLU_OUT1

In reply to Sasha Levin

From: JC Kuo <[email protected]>

[ Upstream commit 0d34dfbf3023cf119b83f6470692c0b10c832495 ]

Full-speed and low-speed USB devices do not work with Tegra210 platforms because of incorrect PLLU/PLLU_OUT1 clock settings.

When full-speed device is connected:

[   14.059886] usb 1-3: new full-speed USB device number 2 using tegra-xusb
[   14.196295] usb 1-3: device descriptor read/64, error -71
[   14.436311] usb 1-3: device descriptor read/64, error -71
[   14.675749] usb 1-3: new full-speed USB device number 3 using tegra-xusb
[   14.812335] usb 1-3: device descriptor read/64, error -71
[   15.052316] usb 1-3: device descriptor read/64, error -71
[   15.164799] usb usb1-port3: attempt power cycle

When low-speed device is connected:

[   37.610949] usb usb1-port3: Cannot enable. Maybe the USB cable is bad?
[   38.557376] usb usb1-port3: Cannot enable. Maybe the USB cable is bad?
[   38.564977] usb usb1-port3: attempt power cycle

This commit fixes the issue by:
1. initializing PLLU_OUT1 before initializing XUSB_FS_SRC clock because PLLU_OUT1 is parent of XUSB_FS_SRC.
2. changing PLLU post-divider to /2 (DIVP=1) according to Technical Reference Manual.

Fixes e745f992cf4b ("clk: tegra: Rework pll_u")
---
 drivers/clk/tegra/clk-tegra210.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/clk/tegra/clk-tegra210.c +4 -4
index 9eb1cb14fce1..4e1bc23c9865 100644
@@ -2214,9 +2214,9 @@ static struct div_nmp pllu_nmp = {
 };
 
 static struct tegra_clk_pll_freq_table pll_u_freq_table[] = {
-	{ 12000000, 480000000, 40, 1, 0, 0 },
-	{ 13000000, 480000000, 36, 1, 0, 0 }, /* actual: 468.0 MHz */
-	{ 38400000, 480000000, 25, 2, 0, 0 },
+	{ 12000000, 480000000, 40, 1, 1, 0 },
+	{ 13000000, 480000000, 36, 1, 1, 0 }, /* actual: 468.0 MHz */
+	{ 38400000, 480000000, 25, 2, 1, 0 },
 	{        0,         0,  0, 0, 0, 0 },
 };
 
@@ -3343,6 +3343,7 @@ static struct tegra_clk_init_table init_table[] __initdata = {
 	{ TEGRA210_CLK_DFLL_REF, TEGRA210_CLK_PLL_P, 51000000, 1 },
 	{ TEGRA210_CLK_SBC4, TEGRA210_CLK_PLL_P, 12000000, 1 },
 	{ TEGRA210_CLK_PLL_RE_VCO, TEGRA210_CLK_CLK_MAX, 672000000, 1 },
+	{ TEGRA210_CLK_PLL_U_OUT1, TEGRA210_CLK_CLK_MAX, 48000000, 1 },
 	{ TEGRA210_CLK_XUSB_GATE, TEGRA210_CLK_CLK_MAX, 0, 1 },
 	{ TEGRA210_CLK_XUSB_SS_SRC, TEGRA210_CLK_PLL_U_480M, 120000000, 0 },
 	{ TEGRA210_CLK_XUSB_FS_SRC, TEGRA210_CLK_PLL_U_48M, 48000000, 0 },
@@ -3367,7 +3368,6 @@ static struct tegra_clk_init_table init_table[] __initdata = {
 	{ TEGRA210_CLK_PLL_DP, TEGRA210_CLK_CLK_MAX, 270000000, 0 },
 	{ TEGRA210_CLK_SOC_THERM, TEGRA210_CLK_PLL_P, 51000000, 0 },
 	{ TEGRA210_CLK_CCLK_G, TEGRA210_CLK_CLK_MAX, 0, 1 },
-	{ TEGRA210_CLK_PLL_U_OUT1, TEGRA210_CLK_CLK_MAX, 48000000, 1 },
 	{ TEGRA210_CLK_PLL_U_OUT2, TEGRA210_CLK_CLK_MAX, 60000000, 1 },
 	/* This MUST be the last entry. */
 	{ TEGRA210_CLK_CLK_MAX, TEGRA210_CLK_CLK_MAX, 0, 0 },
Signature · 1 line
2.20.1
author maintainer
#12 lore raw

PATCH 13/47 fs/adfs: super: fix use-after-free bug

In reply to Sasha Levin

From: Russell King <[email protected]>

[ Upstream commit 5808b14a1f52554de612fee85ef517199855e310 ]

Fix a use-after-free bug during filesystem initialisation, where we access the disc record (which is stored in a buffer) after we have released the buffer.

---
 fs/adfs/super.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
fs/adfs/super.c +4 -1
index 7e099a7a4eb1..4dc15b263489 100644
@@ -369,6 +369,7 @@ static int adfs_fill_super(struct super_block *sb, void *data, int silent)
 	struct buffer_head *bh;
 	struct object_info root_obj;
 	unsigned char *b_data;
+	unsigned int blocksize;
 	struct adfs_sb_info *asb;
 	struct inode *root;
 	int ret = -EINVAL;
@@ -420,8 +421,10 @@ static int adfs_fill_super(struct super_block *sb, void *data, int silent)
 		goto error_free_bh;
 	}
 
+	blocksize = 1 << dr->log2secsize;
 	brelse(bh);
-	if (sb_set_blocksize(sb, 1 << dr->log2secsize)) {
+
+	if (sb_set_blocksize(sb, blocksize)) {
 		bh = sb_bread(sb, ADFS_DISCRECORD / sb->s_blocksize);
 		if (!bh) {
 			adfs_error(sb, "couldn't read superblock on "
Signature · 1 line
2.20.1
author maintainer
#13 lore raw

PATCH 14/47 clk: sprd: Add check for return value of sprd_clk_regmap_init()

In reply to Sasha Levin

From: Chunyan Zhang <[email protected]>

[ Upstream commit c974c48deeb969c5e4250e4f06af91edd84b1f10 ]

sprd_clk_regmap_init() doesn't always return success, adding check for its return value should make the code more strong.

---
 drivers/clk/sprd/sc9860-clk.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
drivers/clk/sprd/sc9860-clk.c +4 -1
index 9980ab55271b..f76305b4bc8d 100644
@@ -2023,6 +2023,7 @@ static int sc9860_clk_probe(struct platform_device *pdev)
 {
 	const struct of_device_id *match;
 	const struct sprd_clk_desc *desc;
+	int ret;
 
 	match = of_match_node(sprd_sc9860_clk_ids, pdev->dev.of_node);
 	if (!match) {
@@ -2031,7 +2032,9 @@ static int sc9860_clk_probe(struct platform_device *pdev)
 	}
 
 	desc = match->data;
-	sprd_clk_regmap_init(pdev, desc);
+	ret = sprd_clk_regmap_init(pdev, desc);
+	if (ret)
+		return ret;
 
 	return sprd_clk_probe(&pdev->dev, desc->hw_clks);
 }
Signature · 1 line
2.20.1
author maintainer
#14 lore raw

PATCH 15/47 btrfs: fix minimum number of chunk errors for DUP

In reply to Sasha Levin

From: David Sterba <[email protected]>

[ Upstream commit 0ee5f8ae082e1f675a2fb6db601c31ac9958a134 ]

The list of profiles in btrfs_chunk_max_errors lists DUP as a profile DUP able to tolerate 1 device missing. Though this profile is special with 2 copies, it still needs the device, unlike the others.

Looking at the history of changes, thre's no clear reason why DUP is there, functions were refactored and blocks of code merged to one helper.

d20983b40e828 Btrfs: fix writing data into the seed filesystem
- factor code to a helper

de11cc12df173 Btrfs: don't pre-allocate btrfs bio
- unrelated change, DUP still in the list with max errors 1

a236aed14ccb0 Btrfs: Deal with failed writes in mirrored configurations
- introduced the max errors, leaves DUP and RAID1 in the same group

---
 fs/btrfs/volumes.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/btrfs/volumes.c +1 -2
index 2fd000308be7..6e008bd5c8cd 100644
@@ -5040,8 +5040,7 @@ static inline int btrfs_chunk_max_errors(struct map_lookup *map)
 
 	if (map->type & (BTRFS_BLOCK_GROUP_RAID1 |
 			 BTRFS_BLOCK_GROUP_RAID10 |
-			 BTRFS_BLOCK_GROUP_RAID5 |
-			 BTRFS_BLOCK_GROUP_DUP)) {
+			 BTRFS_BLOCK_GROUP_RAID5)) {
 		max_errors = 1;
 	} else if (map->type & BTRFS_BLOCK_GROUP_RAID6) {
 		max_errors = 2;
Signature · 1 line
2.20.1
author maintainer
#15 lore raw

PATCH 16/47 btrfs: qgroup: Don't hold qgroup_ioctl_lock in btrfs_qgroup_inherit()

In reply to Sasha Levin

From: Qu Wenruo <[email protected]>

[ Upstream commit e88439debd0a7f969b3ddba6f147152cd0732676 ]

[BUG]
Lockdep will report the following circular locking dependency:

WARNING: possible circular locking dependency detected
5.2.0-rc2-custom #24 Tainted: G O
------------------------------------------------------
btrfs/8631 is trying to acquire lock:
000000002536438c (&fs_info->qgroup_ioctl_lock#2){+.+.}, at: btrfs_qgroup_inherit+0x40/0x620 [btrfs]

but task is already holding lock:
000000003d52cc23 (&fs_info->tree_log_mutex){+.+.}, at: create_pending_snapshot+0x8b6/0xe60 [btrfs]

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #2 (&fs_info->tree_log_mutex){+.+.}:

         __mutex_lock+0x76/0x940
         mutex_lock_nested+0x1b/0x20
         btrfs_commit_transaction+0x475/0xa00 [btrfs]
         btrfs_commit_super+0x71/0x80 [btrfs]
         close_ctree+0x2bd/0x320 [btrfs]
         btrfs_put_super+0x15/0x20 [btrfs]
         generic_shutdown_super+0x72/0x110
         kill_anon_super+0x18/0x30
         btrfs_kill_super+0x16/0xa0 [btrfs]
         deactivate_locked_super+0x3a/0x80
         deactivate_super+0x51/0x60
         cleanup_mnt+0x3f/0x80
         __cleanup_mnt+0x12/0x20
         task_work_run+0x94/0xb0
         exit_to_usermode_loop+0xd8/0xe0
         do_syscall_64+0x210/0x240
         entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #1 (&fs_info->reloc_mutex){+.+.}:

         __mutex_lock+0x76/0x940
         mutex_lock_nested+0x1b/0x20
         btrfs_commit_transaction+0x40d/0xa00 [btrfs]
         btrfs_quota_enable+0x2da/0x730 [btrfs]
         btrfs_ioctl+0x2691/0x2b40 [btrfs]
         do_vfs_ioctl+0xa9/0x6d0
         ksys_ioctl+0x67/0x90
         __x64_sys_ioctl+0x1a/0x20
         do_syscall_64+0x65/0x240
         entry_SYSCALL_64_after_hwframe+0x49/0xbe

-> #0 (&fs_info->qgroup_ioctl_lock#2){+.+.}:

         lock_acquire+0xa7/0x190
         __mutex_lock+0x76/0x940
         mutex_lock_nested+0x1b/0x20
         btrfs_qgroup_inherit+0x40/0x620 [btrfs]
         create_pending_snapshot+0x9d7/0xe60 [btrfs]
         create_pending_snapshots+0x94/0xb0 [btrfs]
         btrfs_commit_transaction+0x415/0xa00 [btrfs]
         btrfs_mksubvol+0x496/0x4e0 [btrfs]
         btrfs_ioctl_snap_create_transid+0x174/0x180 [btrfs]
         btrfs_ioctl_snap_create_v2+0x11c/0x180 [btrfs]
         btrfs_ioctl+0xa90/0x2b40 [btrfs]
         do_vfs_ioctl+0xa9/0x6d0
         ksys_ioctl+0x67/0x90
         __x64_sys_ioctl+0x1a/0x20
         do_syscall_64+0x65/0x240
         entry_SYSCALL_64_after_hwframe+0x49/0xbe

other info that might help us debug this:

  Chain exists of:
    &fs_info->qgroup_ioctl_lock#2 --> &fs_info->reloc_mutex --> &fs_info->tree_log_mutex

   Possible unsafe locking scenario:
         CPU0                    CPU1
         ----                    ----
    lock(&fs_info->tree_log_mutex);
                                 lock(&fs_info->reloc_mutex);
                                 lock(&fs_info->tree_log_mutex);
    lock(&fs_info->qgroup_ioctl_lock#2);

*** DEADLOCK ***

6 locks held by btrfs/8631:
#0: 00000000ed8f23f6 (sb_writers#12){.+.+}, at: mnt_want_write_file+0x28/0x60
#1: 000000009fb1597a (&type->i_mutex_dir_key#10/1){+.+.}, at: btrfs_mksubvol+0x70/0x4e0 [btrfs] #2: 0000000088c5ad88 (&fs_info->subvol_sem){++++}, at: btrfs_mksubvol+0x128/0x4e0 [btrfs] #3: 000000009606fc3e (sb_internal#2){.+.+}, at: start_transaction+0x37a/0x520 [btrfs] #4: 00000000f82bbdf5 (&fs_info->reloc_mutex){+.+.}, at: btrfs_commit_transaction+0x40d/0xa00 [btrfs] #5: 000000003d52cc23 (&fs_info->tree_log_mutex){+.+.}, at: create_pending_snapshot+0x8b6/0xe60 [btrfs]

[CAUSE]
Due to the delayed subvolume creation, we need to call btrfs_qgroup_inherit() inside commit transaction code, with a lot of other mutex hold.
This hell of lock chain can lead to above problem.

[FIX]
On the other hand, we don't really need to hold qgroup_ioctl_lock if we're in the context of create_pending_snapshot().
As in that context, we're the only one being able to modify qgroup.

All other qgroup functions which needs qgroup_ioctl_lock are either holding a transaction handle, or will start a new transaction:
Functions will start a new transaction():
* btrfs_quota_enable()
* btrfs_quota_disable()
Functions hold a transaction handler:
* btrfs_add_qgroup_relation()
* btrfs_del_qgroup_relation()
* btrfs_create_qgroup()
* btrfs_remove_qgroup()
* btrfs_limit_qgroup()
* btrfs_qgroup_inherit() call inside create_subvol()

So we have a higher level protection provided by transaction, thus we don't need to always hold qgroup_ioctl_lock in btrfs_qgroup_inherit().

Only the btrfs_qgroup_inherit() call in create_subvol() needs to hold qgroup_ioctl_lock, while the btrfs_qgroup_inherit() call in create_pending_snapshot() is already protected by transaction.

So the fix is to detect the context by checking trans->transaction->state.
If we're at TRANS_STATE_COMMIT_DOING, then we're in commit transaction context and no need to get the mutex.

---
 fs/btrfs/qgroup.c | 24 ++++++++++++++++++++++--
 1 file changed, 22 insertions(+), 2 deletions(-)
fs/btrfs/qgroup.c +22 -2
index e46e83e87600..734866ab5194 100644
@@ -2249,6 +2249,7 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, u64 srcid,
 	int ret = 0;
 	int i;
 	u64 *i_qgroups;
+	bool committing = false;
 	struct btrfs_fs_info *fs_info = trans->fs_info;
 	struct btrfs_root *quota_root;
 	struct btrfs_qgroup *srcgroup;
@@ -2256,7 +2257,25 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, u64 srcid,
 	u32 level_size = 0;
 	u64 nums;
 
-	mutex_lock(&fs_info->qgroup_ioctl_lock);
+	/*
+	 * There are only two callers of this function.
+	 *
+	 * One in create_subvol() in the ioctl context, which needs to hold
+	 * the qgroup_ioctl_lock.
+	 *
+	 * The other one in create_pending_snapshot() where no other qgroup
+	 * code can modify the fs as they all need to either start a new trans
+	 * or hold a trans handler, thus we don't need to hold
+	 * qgroup_ioctl_lock.
+	 * This would avoid long and complex lock chain and make lockdep happy.
+	 */
+	spin_lock(&fs_info->trans_lock);
+	if (trans->transaction->state == TRANS_STATE_COMMIT_DOING)
+		committing = true;
+	spin_unlock(&fs_info->trans_lock);
+
+	if (!committing)
+		mutex_lock(&fs_info->qgroup_ioctl_lock);
 	if (!test_bit(BTRFS_FS_QUOTA_ENABLED, &fs_info->flags))
 		goto out;
 
@@ -2420,7 +2439,8 @@ int btrfs_qgroup_inherit(struct btrfs_trans_handle *trans, u64 srcid,
 unlock:
 	spin_unlock(&fs_info->qgroup_lock);
 out:
-	mutex_unlock(&fs_info->qgroup_ioctl_lock);
+	if (!committing)
+		mutex_unlock(&fs_info->qgroup_ioctl_lock);
 	return ret;
 }
 
Signature · 1 line
2.20.1
author maintainer
#16 lore raw

PATCH 17/47 cifs: Fix a race condition with cifs_echo_request

In reply to Sasha Levin

From: Ronnie Sahlberg <[email protected]>

[ Upstream commit f2caf901c1b7ce65f9e6aef4217e3241039db768 ]

There is a race condition with how we send (or supress and don't send) smb echos that will cause the client to incorrectly think the server is unresponsive and thus needs to be reconnected.

Summary of the race condition:
1) Daisy chaining scheduling creates a gap.
2) If traffic comes unfortunate shortly after the last echo, the planned echo is suppressed.
3) Due to the gap, the next echo transmission is delayed until after the timeout, which is set hard to twice the echo interval.

This is fixed by changing the timeouts from 2 to three times the echo interval.

Detailed description of the bug: https://lutz.donnerhacke.de/eng/Blog/Groundhog-Day-with-SMB-remount

---
 fs/cifs/connect.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
fs/cifs/connect.c +4 -4
index f31339db45fd..c53a2e86ed54 100644
@@ -563,10 +563,10 @@ static bool
 server_unresponsive(struct TCP_Server_Info *server)
 {
 	/*
-	 * We need to wait 2 echo intervals to make sure we handle such
+	 * We need to wait 3 echo intervals to make sure we handle such
 	 * situations right:
 	 * 1s  client sends a normal SMB request
-	 * 2s  client gets a response
+	 * 3s  client gets a response
 	 * 30s echo workqueue job pops, and decides we got a response recently
 	 *     and don't need to send another
 	 * ...
@@ -575,9 +575,9 @@ server_unresponsive(struct TCP_Server_Info *server)
 	 */
 	if ((server->tcpStatus == CifsGood ||
 	    server->tcpStatus == CifsNeedNegotiate) &&
-	    time_after(jiffies, server->lstrp + 2 * server->echo_interval)) {
+	    time_after(jiffies, server->lstrp + 3 * server->echo_interval)) {
 		cifs_dbg(VFS, "Server %s has not responded in %lu seconds. Reconnecting...\n",
-			 server->hostname, (2 * server->echo_interval) / HZ);
+			 server->hostname, (3 * server->echo_interval) / HZ);
 		cifs_reconnect(server);
 		wake_up(&server->response_q);
 		return true;
Signature · 1 line
2.20.1
author maintainer
#17 lore raw

PATCH 18/47 ceph: fix listxattr vxattr buffer length calculation

In reply to Sasha Levin

From: David Disseldorp <[email protected]>

[ Upstream commit 2b2abcac8c251d1c77a4cc9d9f248daefae0fb4e ]

ceph_listxattr() incorrectly returns a length based on the static ceph_vxattrs_name_size() value, which only takes into account whether vxattrs are hidden, ignoring vxattr.exists_cb().

When filling the xattr buffer ceph_listxattr() checks VXATTR_FLAG_HIDDEN and vxattr.exists_cb(). If both are false, we return an incorrect (oversize) length.

Fix this behaviour by always calculating the vxattrs length at runtime, taking both vxattr.hidden and vxattr.exists_cb() into account.

This bug is only exposed with the new "ceph.snap.btime" vxattr, as all other vxattrs with a non-null exists_cb also carry VXATTR_FLAG_HIDDEN.

---
 fs/ceph/xattr.c | 54 +++++++++++++++++++++++++++----------------------
 1 file changed, 30 insertions(+), 24 deletions(-)
fs/ceph/xattr.c +30 -24
index 5cc8b94f8206..996ee87b1eaf 100644
@@ -879,10 +879,9 @@ ssize_t ceph_listxattr(struct dentry *dentry, char *names, size_t size)
 	struct inode *inode = d_inode(dentry);
 	struct ceph_inode_info *ci = ceph_inode(inode);
 	struct ceph_vxattr *vxattrs = ceph_inode_vxattrs(inode);
-	u32 vir_namelen = 0;
+	bool len_only = (size == 0);
 	u32 namelen;
 	int err;
-	u32 len;
 	int i;
 
 	spin_lock(&ci->i_ceph_lock);
@@ -901,38 +900,45 @@ ssize_t ceph_listxattr(struct dentry *dentry, char *names, size_t size)
 	err = __build_xattrs(inode);
 	if (err < 0)
 		goto out;
-	/*
-	 * Start with virtual dir xattr names (if any) (including
-	 * terminating '\0' characters for each).
-	 */
-	vir_namelen = ceph_vxattrs_name_size(vxattrs);
 
-	/* adding 1 byte per each variable due to the null termination */
+	/* add 1 byte for each xattr due to the null termination */
 	namelen = ci->i_xattrs.names_size + ci->i_xattrs.count;
-	err = -ERANGE;
-	if (size && vir_namelen + namelen > size)
-		goto out;
-
-	err = namelen + vir_namelen;
-	if (size == 0)
-		goto out;
+	if (!len_only) {
+		if (namelen > size) {
+			err = -ERANGE;
+			goto out;
+		}
+		names = __copy_xattr_names(ci, names);
+		size -= namelen;
+	}
 
-	names = __copy_xattr_names(ci, names);
 
 	/* virtual xattr names, too */
-	err = namelen;
 	if (vxattrs) {
 		for (i = 0; vxattrs[i].name; i++) {
-			if (!(vxattrs[i].flags & VXATTR_FLAG_HIDDEN) &&
-			    !(vxattrs[i].exists_cb &&
-			      !vxattrs[i].exists_cb(ci))) {
-				len = sprintf(names, "%s", vxattrs[i].name);
-				names += len + 1;
-				err += len + 1;
+			size_t this_len;
+
+			if (vxattrs[i].flags & VXATTR_FLAG_HIDDEN)
+				continue;
+			if (vxattrs[i].exists_cb && !vxattrs[i].exists_cb(ci))
+				continue;
+
+			this_len = strlen(vxattrs[i].name) + 1;
+			namelen += this_len;
+			if (len_only)
+				continue;
+
+			if (this_len > size) {
+				err = -ERANGE;
+				goto out;
 			}
+
+			memcpy(names, vxattrs[i].name, this_len);
+			names += this_len;
+			size -= this_len;
 		}
 	}
-
+	err = namelen;
 out:
 	spin_unlock(&ci->i_ceph_lock);
 	return err;
Signature · 1 line
2.20.1
author maintainer
#18 lore raw

PATCH 19/47 ceph: fix improper use of smp_mb__before_atomic()

In reply to Sasha Levin

From: Andrea Parri <[email protected]>

[ Upstream commit 749607731e26dfb2558118038c40e9c0c80d23b5 ]

This barrier only applies to the read-modify-write operations; in particular, it does not apply to the atomic64_set() primitive.

Replace the barrier with an smp_mb().

---
 fs/ceph/super.h | 7 ++++++-
 1 file changed, 6 insertions(+), 1 deletion(-)
fs/ceph/super.h +6 -1
index 582e28fd1b7b..d8579a56e5dc 100644
@@ -526,7 +526,12 @@ static inline void __ceph_dir_set_complete(struct ceph_inode_info *ci,
 					   long long release_count,
 					   long long ordered_count)
 {
-	smp_mb__before_atomic();
+	/*
+	 * Makes sure operations that setup readdir cache (update page
+	 * cache and i_size) are strongly ordered w.r.t. the following
+	 * atomic64_set() operations.
+	 */
+	smp_mb();
 	atomic64_set(&ci->i_complete_seq[0], release_count);
 	atomic64_set(&ci->i_complete_seq[1], ordered_count);
 }
Signature · 1 line
2.20.1
author maintainer
#19 lore raw

PATCH 20/47 ceph: return -ERANGE if virtual xattr value didn't fit in buffer

In reply to Sasha Levin

From: Jeff Layton <[email protected]>

[ Upstream commit 3b421018f48c482bdc9650f894aa1747cf90e51d ]

The getxattr manpage states that we should return ERANGE if the destination buffer size is too small to hold the value. ceph_vxattrcb_layout does this internally, but we should be doing this for all vxattrs.

Fix the only caller of getxattr_cb to check the returned size against the buffer length and return -ERANGE if it doesn't fit. Drop the same check in ceph_vxattrcb_layout and just rely on the caller to handle it.

---
 fs/ceph/xattr.c | 14 +++++++-------
 1 file changed, 7 insertions(+), 7 deletions(-)
fs/ceph/xattr.c +7 -7
index 996ee87b1eaf..7e6d3df99f2f 100644
@@ -79,7 +79,7 @@ static size_t ceph_vxattrcb_layout(struct ceph_inode_info *ci, char *val,
 	const char *ns_field = " pool_namespace=";
 	char buf[128];
 	size_t len, total_len = 0;
-	int ret;
+	ssize_t ret;
 
 	pool_ns = ceph_try_get_string(ci->i_layout.pool_ns);
 
@@ -103,11 +103,8 @@ static size_t ceph_vxattrcb_layout(struct ceph_inode_info *ci, char *val,
 	if (pool_ns)
 		total_len += strlen(ns_field) + pool_ns->len;
 
-	if (!size) {
-		ret = total_len;
-	} else if (total_len > size) {
-		ret = -ERANGE;
-	} else {
+	ret = total_len;
+	if (size >= total_len) {
 		memcpy(val, buf, len);
 		ret = len;
 		if (pool_name) {
@@ -817,8 +814,11 @@ ssize_t __ceph_getxattr(struct inode *inode, const char *name, void *value,
 		if (err)
 			return err;
 		err = -ENODATA;
-		if (!(vxattr->exists_cb && !vxattr->exists_cb(ci)))
+		if (!(vxattr->exists_cb && !vxattr->exists_cb(ci))) {
 			err = vxattr->getxattr_cb(ci, value, size);
+			if (size && size < err)
+				err = -ERANGE;
+		}
 		return err;
 	}
 
Signature · 1 line
2.20.1
author maintainer
#20 lore raw

PATCH 21/47 ACPI: blacklist: fix clang warning for unused DMI table

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit b80d6a42bdc97bdb6139107d6034222e9843c6e2 ]

When CONFIG_DMI is disabled, we only have a tentative declaration, which causes a warning from clang:

drivers/acpi/blacklist.c:20:35: error: tentative array definition assumed to have one element [-Werror] static const struct dmi_system_id acpi_rev_dmi_table[] __initconst;

As the variable is not actually used here, hide it entirely in an #ifdef to shut up the warning.

---
 drivers/acpi/blacklist.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/acpi/blacklist.c +4 -0
index 995c4d8922b1..761f0c19a451 100644
@@ -30,7 +30,9 @@
 
 #include "internal.h"
 
+#ifdef CONFIG_DMI
 static const struct dmi_system_id acpi_rev_dmi_table[] __initconst;
+#endif
 
 /*
  * POLICY: If *anything* doesn't work, put it on the blacklist.
@@ -74,7 +76,9 @@ int __init acpi_blacklisted(void)
 	}
 
 	(void)early_acpi_osi_init();
+#ifdef CONFIG_DMI
 	dmi_check_system(acpi_rev_dmi_table);
+#endif
 
 	return blacklisted;
 }
Signature · 1 line
2.20.1
author maintainer
#21 lore raw

PATCH 22/47 scsi: zfcp: fix GCC compiler warning emitted with -Wmaybe-uninitialized

In reply to Sasha Levin

From: Benjamin Block <[email protected]>

[ Upstream commit 484647088826f2f651acbda6bcf9536b8a466703 ]

GCC v9 emits this warning:
CC drivers/s390/scsi/zfcp_erp.o
drivers/s390/scsi/zfcp_erp.c: In function 'zfcp_erp_action_enqueue':
drivers/s390/scsi/zfcp_erp.c:217:26: warning: 'erp_action' may be used uninitialized in this function [-Wmaybe-uninitialized]
217 | struct zfcp_erp_action *erp_action;
| ^~~~~~~~~~

This is a possible false positive case, as also documented in the GCC documentations:
https://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html#index-Wmaybe-uninitialized

The actual code-sequence is like this:

    Various callers can invoke the function below with the argument "want"
    being one of:
    ZFCP_ERP_ACTION_REOPEN_ADAPTER,
    ZFCP_ERP_ACTION_REOPEN_PORT_FORCED,
    ZFCP_ERP_ACTION_REOPEN_PORT, or
    ZFCP_ERP_ACTION_REOPEN_LUN.

    zfcp_erp_action_enqueue(want, ...)
        ...
        need = zfcp_erp_required_act(want, ...)
            need = want
            ...
            maybe: need = ZFCP_ERP_ACTION_REOPEN_PORT
            maybe: need = ZFCP_ERP_ACTION_REOPEN_ADAPTER
            ...
            return need
        ...
        zfcp_erp_setup_act(need, ...)
            struct zfcp_erp_action *erp_action; // <== line 217
            ...
            switch(need) {
            case ZFCP_ERP_ACTION_REOPEN_LUN:
                    ...
                    erp_action = &zfcp_sdev->erp_action;
                    WARN_ON_ONCE(erp_action->port != port); // <== access
                    ...
                    break;
            case ZFCP_ERP_ACTION_REOPEN_PORT:
            case ZFCP_ERP_ACTION_REOPEN_PORT_FORCED:
                    ...
                    erp_action = &port->erp_action;
                    WARN_ON_ONCE(erp_action->port != port); // <== access
                    ...
                    break;
            case ZFCP_ERP_ACTION_REOPEN_ADAPTER:
                    ...
                    erp_action = &adapter->erp_action;
                    WARN_ON_ONCE(erp_action->port != NULL); // <== access
                    ...
                    break;
            }
            ...
            WARN_ON_ONCE(erp_action->adapter != adapter); // <== access

When zfcp_erp_setup_act() is called, 'need' will never be anything else than one of the 4 possible enumeration-names that are used in the switch-case, and 'erp_action' is initialized for every one of them, before it is used. Thus the warning is a false positive, as documented.

We introduce the extra if{} in the beginning to create an extra code-flow, so the compiler can be convinced that the switch-case will never see any other value.

BUG_ON()/BUG() is intentionally not used to not crash anything, should this ever happen anyway - right now it's impossible, as argued above; and it doesn't introduce a 'default:' switch-case to retain warnings should 'enum zfcp_erp_act_type' ever be extended and no explicit case be introduced. See also v5.0 commit 399b6c8bc9f7 ("scsi: zfcp: drop old default switch case which might paper over missing case").

---
 drivers/s390/scsi/zfcp_erp.c | 7 +++++++
 1 file changed, 7 insertions(+)
drivers/s390/scsi/zfcp_erp.c +7 -0
index ebdbc457003f..332701db7379 100644
@@ -11,6 +11,7 @@
 #define pr_fmt(fmt) KMSG_COMPONENT ": " fmt
 
 #include <linux/kthread.h>
+#include <linux/bug.h>
 #include "zfcp_ext.h"
 #include "zfcp_reqlist.h"
 
@@ -238,6 +239,12 @@ static struct zfcp_erp_action *zfcp_erp_setup_act(int need, u32 act_status,
 	struct zfcp_erp_action *erp_action;
 	struct zfcp_scsi_dev *zfcp_sdev;
 
+	if (WARN_ON_ONCE(need != ZFCP_ERP_ACTION_REOPEN_LUN &&
+			 need != ZFCP_ERP_ACTION_REOPEN_PORT &&
+			 need != ZFCP_ERP_ACTION_REOPEN_PORT_FORCED &&
+			 need != ZFCP_ERP_ACTION_REOPEN_ADAPTER))
+		return NULL;
+
 	switch (need) {
 	case ZFCP_ERP_ACTION_REOPEN_LUN:
 		zfcp_sdev = sdev_to_zfcp(sdev);
Signature · 1 line
2.20.1
author maintainer
#22 lore raw

PATCH 23/47 perf version: Fix segfault due to missing OPT_END()

In reply to Sasha Levin

From: Ravi Bangoria <[email protected]>

[ Upstream commit 916c31fff946fae0e05862f9b2435fdb29fd5090 ]

'perf version' on powerpc segfaults when used with non-supported option:

  # perf version -a
  Segmentation fault (core dumped)

Fix this.

---
 tools/perf/builtin-version.c | 1 +
 1 file changed, 1 insertion(+)
tools/perf/builtin-version.c +1 -0
index 50df168be326..b02c96104640 100644
@@ -19,6 +19,7 @@ static struct version version;
 static struct option version_options[] = {
 	OPT_BOOLEAN(0, "build-options", &version.build_options,
 		    "display the build options"),
+	OPT_END(),
 };
 
 static const char * const version_usage[] = {
Signature · 1 line
2.20.1
author maintainer
#23 lore raw

PATCH 24/47 x86: kvm: avoid constant-conversion warning

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit a6a6d3b1f867d34ba5bd61aa7bb056b48ca67cff ]

clang finds a contruct suspicious that converts an unsigned character to a signed integer and back, causing an overflow:

arch/x86/kvm/mmu.c:4605:39: error: implicit conversion from 'int' to 'u8' (aka 'unsigned char') changes value from -205 to 51 [-Werror,-Wconstant-conversion]

                u8 wf = (pfec & PFERR_WRITE_MASK) ? ~w : 0;
                   ~~                               ^~
arch/x86/kvm/mmu.c:4607:38: error: implicit conversion from 'int' to 'u8' (aka 'unsigned char') changes value from -241 to 15 [-Werror,-Wconstant-conversion]
                u8 uf = (pfec & PFERR_USER_MASK) ? ~u : 0;
                   ~~                              ^~
arch/x86/kvm/mmu.c:4609:39: error: implicit conversion from 'int' to 'u8' (aka 'unsigned char') changes value from -171 to 85 [-Werror,-Wconstant-conversion]
                u8 ff = (pfec & PFERR_FETCH_MASK) ? ~x : 0;
                   ~~                               ^~

Add an explicit cast to tell clang that everything works as intended here.

---
 arch/x86/kvm/mmu.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
arch/x86/kvm/mmu.c +3 -3
index e0f982e35c96..cdc0c460950f 100644
@@ -4532,11 +4532,11 @@ static void update_permission_bitmask(struct kvm_vcpu *vcpu,
 		 */
 
 		/* Faults from writes to non-writable pages */
-		u8 wf = (pfec & PFERR_WRITE_MASK) ? ~w : 0;
+		u8 wf = (pfec & PFERR_WRITE_MASK) ? (u8)~w : 0;
 		/* Faults from user mode accesses to supervisor pages */
-		u8 uf = (pfec & PFERR_USER_MASK) ? ~u : 0;
+		u8 uf = (pfec & PFERR_USER_MASK) ? (u8)~u : 0;
 		/* Faults from fetches of non-executable pages*/
-		u8 ff = (pfec & PFERR_FETCH_MASK) ? ~x : 0;
+		u8 ff = (pfec & PFERR_FETCH_MASK) ? (u8)~x : 0;
 		/* Faults from kernel mode fetches of user pages */
 		u8 smepf = 0;
 		/* Faults from kernel mode accesses of user pages */
Signature · 1 line
2.20.1
author maintainer
#24 lore raw

PATCH 25/47 ACPI: fix false-positive -Wuninitialized warning

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit dfd6f9ad36368b8dbd5f5a2b2f0a4705ae69a323 ]

clang gets confused by an uninitialized variable in what looks to it like a never executed code path:

arch/x86/kernel/acpi/boot.c:618:13: error: variable 'polarity' is uninitialized when used here [-Werror,-Wuninitialized]

        polarity = polarity ? ACPI_ACTIVE_LOW : ACPI_ACTIVE_HIGH;
                   ^~~~~~~~
arch/x86/kernel/acpi/boot.c:606:32: note: initialize the variable 'polarity' to silence this warning
        int rc, irq, trigger, polarity;
                                      ^
                                       = 0
arch/x86/kernel/acpi/boot.c:617:12: error: variable 'trigger' is uninitialized when used here [-Werror,-Wuninitialized]
        trigger = trigger ? ACPI_LEVEL_SENSITIVE : ACPI_EDGE_SENSITIVE;
                  ^~~~~~~
arch/x86/kernel/acpi/boot.c:606:22: note: initialize the variable 'trigger' to silence this warning
        int rc, irq, trigger, polarity;
                            ^
                             = 0

This is unfortunately a design decision in clang and won't be fixed.

Changing the acpi_get_override_irq() macro to an inline function reliably avoids the issue.

---
 include/linux/acpi.h | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
include/linux/acpi.h +4 -1
index de8d3d3fa651..b4d23b3a2ef2 100644
@@ -326,7 +326,10 @@ void acpi_set_irq_model(enum acpi_irq_model_id model,
 #ifdef CONFIG_X86_IO_APIC
 extern int acpi_get_override_irq(u32 gsi, int *trigger, int *polarity);
 #else
-#define acpi_get_override_irq(gsi, trigger, polarity) (-1)
+static inline int acpi_get_override_irq(u32 gsi, int *trigger, int *polarity)
+{
+	return -1;
+}
 #endif
 /*
  * This function undoes the effect of one call to acpi_register_gsi().
Signature · 1 line
2.20.1
author maintainer
#25 lore raw

PATCH 26/47 ISDN: hfcsusb: checking idx of ep configuration

In reply to Sasha Levin

From: Phong Tran <[email protected]>

[ Upstream commit f384e62a82ba5d85408405fdd6aeff89354deaa9 ]

The syzbot test with random endpoint address which made the idx is overflow in the table of endpoint configuations.

this adds the checking for fixing the error report from syzbot

KASAN: stack-out-of-bounds Read in hfcsusb_probe [1]

The patch tested by syzbot [2]

Reported-by [email protected]

[1]:
https://syzkaller.appspot.com/bug?id=30a04378dac680c5d521304a00a86156bb913522
[2]:
https://groups.google.com/d/msg/syzkaller-bugs/_6HBdge8F3E/OJn7wVNpBAAJ

---
 drivers/isdn/hardware/mISDN/hfcsusb.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/isdn/hardware/mISDN/hfcsusb.c +3 -0
index 6d05946b445e..060dc7fd66c1 100644
@@ -1967,6 +1967,9 @@ hfcsusb_probe(struct usb_interface *intf, const struct usb_device_id *id)
 
 				/* get endpoint base */
 				idx = ((ep_addr & 0x7f) - 1) * 2;
+				if (idx > 15)
+					return -EIO;
+
 				if (ep_addr & 0x80)
 					idx++;
 				attr = ep->desc.bmAttributes;
Signature · 1 line
2.20.1
author maintainer
#26 lore raw

PATCH 27/47 be2net: Signal that the device cannot transmit during reconfiguration

In reply to Sasha Levin

From: Benjamin Poirier <[email protected]>

[ Upstream commit 7429c6c0d9cb086d8e79f0d2a48ae14851d2115e ]

While changing the number of interrupt channels, be2net stops adapter operation (including netif_tx_disable()) but it doesn't signal that it cannot transmit. This may lead dev_watchdog() to falsely trigger during that time.

Add the missing call to netif_carrier_off(), following the pattern used in many other drivers. netif_carrier_on() is already taken care of in be_open().

---
 drivers/net/ethernet/emulex/benet/be_main.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/net/ethernet/emulex/benet/be_main.c +5 -1
index bff74752cef1..3fe6a28027fe 100644
@@ -4700,8 +4700,12 @@ int be_update_queues(struct be_adapter *adapter)
 	struct net_device *netdev = adapter->netdev;
 	int status;
 
-	if (netif_running(netdev))
+	if (netif_running(netdev)) {
+		/* device cannot transmit now, avoid dev_watchdog timeouts */
+		netif_carrier_off(netdev);
+
 		be_close(netdev);
+	}
 
 	be_cancel_worker(adapter);
 
Signature · 1 line
2.20.1
author maintainer
#27 lore raw

PATCH 28/47 x86/apic: Silence -Wtype-limits compiler warnings

In reply to Sasha Levin

From: Qian Cai <[email protected]>

[ Upstream commit ec6335586953b0df32f83ef696002063090c7aef ]

There are many compiler warnings like this,

In file included from ./arch/x86/include/asm/smp.h:13,

                 from ./arch/x86/include/asm/mmzone_64.h:11,
                 from ./arch/x86/include/asm/mmzone.h:5,
                 from ./include/linux/mmzone.h:969,
                 from ./include/linux/gfp.h:6,
                 from ./include/linux/mm.h:10,
                 from arch/x86/kernel/apic/io_apic.c:34:
arch/x86/kernel/apic/io_apic.c: In function 'check_timer':
./arch/x86/include/asm/apic.h:37:11: warning: comparison of unsigned
expression >= 0 is always true [-Wtype-limits]
   if ((v) <= apic_verbosity) \
           ^~

arch/x86/kernel/apic/io_apic.c:2160:2: note: in expansion of macro 'apic_printk'
apic_printk(APIC_QUIET, KERN_INFO "..TIMER: vector=0x%02X " ^~~~~~~~~~~
./arch/x86/include/asm/apic.h:37:11: warning: comparison of unsigned expression >= 0 is always true [-Wtype-limits]

   if ((v) <= apic_verbosity) \
           ^~

arch/x86/kernel/apic/io_apic.c:2207:4: note: in expansion of macro 'apic_printk'
apic_printk(APIC_QUIET, KERN_ERR "..MP-BIOS bug: " ^~~~~~~~~~~

APIC_QUIET is 0, so silence them by making apic_verbosity type int.

---
 arch/x86/include/asm/apic.h | 2 +-
 arch/x86/kernel/apic/apic.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
arch/x86/include/asm/apic.h +1 -1
index 130e81e10fc7..050368db9d35 100644
@@ -48,7 +48,7 @@ static inline void generic_apic_probe(void)
 
 #ifdef CONFIG_X86_LOCAL_APIC
 
-extern unsigned int apic_verbosity;
+extern int apic_verbosity;
 extern int local_apic_timer_c2_ok;
 
 extern int disable_apic;
arch/x86/kernel/apic/apic.c +1 -1
index 02020f2e0080..272a12865b2a 100644
@@ -181,7 +181,7 @@ EXPORT_SYMBOL_GPL(local_apic_timer_c2_ok);
 /*
  * Debug level, exported for io_apic.c
  */
-unsigned int apic_verbosity;
+int apic_verbosity;
 
 int pic_mode;
 
Signature · 1 line
2.20.1
author maintainer
#28 lore raw

PATCH 29/47 x86: math-emu: Hide clang warnings for 16-bit overflow

In reply to Sasha Levin

From: Arnd Bergmann <[email protected]>

[ Upstream commit 29e7e9664aec17b94a9c8c5a75f8d216a206aa3a ]

clang warns about a few parts of the math-emu implementation where a 16-bit integer becomes negative during assignment:

arch/x86/math-emu/poly_tan.c:88:35: error: implicit conversion from 'int' to 'short' changes value from 49216 to -16320 [-Werror,-Wconstant-conversion]

                                      (0x41 + EXTENDED_Ebias) | SIGN_Negative);
                                      ~~~~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~
arch/x86/math-emu/fpu_emu.h:180:58: note: expanded from macro 'setexponent16'
 #define setexponent16(x,y)  { (*(short *)&((x)->exp)) = (y); }
                                                      ~  ^
arch/x86/math-emu/reg_constant.c:37:32: error: implicit conversion from 'int' to 'short' changes value from 49085 to -16451 [-Werror,-Wconstant-conversion]
FPU_REG const CONST_PI2extra = MAKE_REG(NEG, -66,
                               ^~~~~~~~~~~~~~~~~~
arch/x86/math-emu/reg_constant.c:21:25: note: expanded from macro 'MAKE_REG'
                ((EXTENDED_Ebias+(e)) | ((SIGN_##s != 0)*0x8000)) }
                 ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~
arch/x86/math-emu/reg_constant.c:48:28: error: implicit conversion from 'int' to 'short' changes value from 65535 to -1 [-Werror,-Wconstant-conversion]
FPU_REG const CONST_QNaN = MAKE_REG(NEG, EXP_OVER, 0x00000000, 0xC0000000);
                           ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
arch/x86/math-emu/reg_constant.c:21:25: note: expanded from macro 'MAKE_REG'
                ((EXTENDED_Ebias+(e)) | ((SIGN_##s != 0)*0x8000)) }
                 ~~~~~~~~~~~~~~~~~~~~~^~~~~~~~~~~~~~~~~~~~~~~~~~

The code is correct as is, so add a typecast to shut up the warnings.

---
 arch/x86/math-emu/fpu_emu.h      | 2 +-
 arch/x86/math-emu/reg_constant.c | 2 +-
 2 files changed, 2 insertions(+), 2 deletions(-)
arch/x86/math-emu/fpu_emu.h +1 -1
index a5a41ec58072..0c122226ca56 100644
@@ -177,7 +177,7 @@ static inline void reg_copy(FPU_REG const *x, FPU_REG *y)
 #define setexponentpos(x,y) { (*(short *)&((x)->exp)) = \
   ((y) + EXTENDED_Ebias) & 0x7fff; }
 #define exponent16(x)         (*(short *)&((x)->exp))
-#define setexponent16(x,y)  { (*(short *)&((x)->exp)) = (y); }
+#define setexponent16(x,y)  { (*(short *)&((x)->exp)) = (u16)(y); }
 #define addexponent(x,y)    { (*(short *)&((x)->exp)) += (y); }
 #define stdexp(x)           { (*(short *)&((x)->exp)) += EXTENDED_Ebias; }
 
arch/x86/math-emu/reg_constant.c +1 -1
index 8dc9095bab22..742619e94bdf 100644
@@ -18,7 +18,7 @@
 #include "control_w.h"
 
 #define MAKE_REG(s, e, l, h) { l, h, \
-		((EXTENDED_Ebias+(e)) | ((SIGN_##s != 0)*0x8000)) }
+		(u16)((EXTENDED_Ebias+(e)) | ((SIGN_##s != 0)*0x8000)) }
 
 FPU_REG const CONST_1 = MAKE_REG(POS, 0, 0x00000000, 0x80000000);
 #if 0
Signature · 1 line
2.20.1
author maintainer
#29 lore raw

PATCH 30/47 mm/cma.c: fail if fixed declaration can't be honored

In reply to Sasha Levin

From: Doug Berger <[email protected]>

[ Upstream commit c633324e311243586675e732249339685e5d6faa ]

The description of cma_declare_contiguous() indicates that if the 'fixed' argument is true the reserved contiguous area must be exactly at the address of the 'base' argument.

However, the function currently allows the 'base', 'size', and 'limit' arguments to be silently adjusted to meet alignment constraints. This commit enforces the documented behavior through explicit checks that return an error if the region does not fit within a specified region.

Fixes 5ea3b1b2f8ad ("cma: add placement specifier for "cma=" kernel parameter")
Cc Yue Hu <[email protected]>
Cc Mike Rapoport <[email protected]>
Cc Laura Abbott <[email protected]>
Cc Peng Fan <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Marek Szyprowski <[email protected]>
Cc Andrey Konovalov <[email protected]>
---
 mm/cma.c | 13 +++++++++++++
 1 file changed, 13 insertions(+)
mm/cma.c +13 -0
index 476dfe13a701..4c2864270a39 100644
@@ -282,6 +282,12 @@ int __init cma_declare_contiguous(phys_addr_t base,
 	 */
 	alignment = max(alignment,  (phys_addr_t)PAGE_SIZE <<
 			  max_t(unsigned long, MAX_ORDER - 1, pageblock_order));
+	if (fixed && base & (alignment - 1)) {
+		ret = -EINVAL;
+		pr_err("Region at %pa must be aligned to %pa bytes\n",
+			&base, &alignment);
+		goto err;
+	}
 	base = ALIGN(base, alignment);
 	size = ALIGN(size, alignment);
 	limit &= ~(alignment - 1);
@@ -312,6 +318,13 @@ int __init cma_declare_contiguous(phys_addr_t base,
 	if (limit == 0 || limit > memblock_end)
 		limit = memblock_end;
 
+	if (base + size > limit) {
+		ret = -EINVAL;
+		pr_err("Size (%pa) of region at %pa exceeds limit (%pa)\n",
+			&size, &base, &limit);
+		goto err;
+	}
+
 	/* Reserve memory */
 	if (fixed) {
 		if (memblock_is_region_reserved(base, size) ||
Signature · 1 line
2.20.1
author maintainer
#30 lore raw

PATCH 31/47 lib/test_overflow.c: avoid tainting the kernel and fix wrap size

In reply to Sasha Levin

From: Kees Cook <[email protected]>

[ Upstream commit 8e060c21ae2c265a2b596e9e7f9f97ec274151a4 ]

This adds __GFP_NOWARN to the kmalloc()-portions of the overflow test to avoid tainting the kernel. Additionally fixes up the math on wrap size to be architecture and page size agnostic.

---
 lib/test_overflow.c | 11 ++++++-----
 1 file changed, 6 insertions(+), 5 deletions(-)
lib/test_overflow.c +6 -5
index fc680562d8b6..7a4b6f6c5473 100644
@@ -486,16 +486,17 @@ static int __init test_overflow_shift(void)
  * Deal with the various forms of allocator arguments. See comments above
  * the DEFINE_TEST_ALLOC() instances for mapping of the "bits".
  */
-#define alloc010(alloc, arg, sz) alloc(sz, GFP_KERNEL)
-#define alloc011(alloc, arg, sz) alloc(sz, GFP_KERNEL, NUMA_NO_NODE)
+#define alloc_GFP		 (GFP_KERNEL | __GFP_NOWARN)
+#define alloc010(alloc, arg, sz) alloc(sz, alloc_GFP)
+#define alloc011(alloc, arg, sz) alloc(sz, alloc_GFP, NUMA_NO_NODE)
 #define alloc000(alloc, arg, sz) alloc(sz)
 #define alloc001(alloc, arg, sz) alloc(sz, NUMA_NO_NODE)
-#define alloc110(alloc, arg, sz) alloc(arg, sz, GFP_KERNEL)
+#define alloc110(alloc, arg, sz) alloc(arg, sz, alloc_GFP)
 #define free0(free, arg, ptr)	 free(ptr)
 #define free1(free, arg, ptr)	 free(arg, ptr)
 
-/* Wrap around to 8K */
-#define TEST_SIZE		(9 << PAGE_SHIFT)
+/* Wrap around to 16K */
+#define TEST_SIZE		(5 * 4096)
 
 #define DEFINE_TEST_ALLOC(func, free_func, want_arg, want_gfp, want_node)\
 static int __init test_ ## func (void *arg)				\
Signature · 1 line
2.20.1
author maintainer
#31 lore raw

PATCH 32/47 lib/test_string.c: avoid masking memset16/32/64 failures

In reply to Sasha Levin

From: Peter Rosin <[email protected]>

[ Upstream commit 33d6e0ff68af74be0c846c8e042e84a9a1a0561e ]

If a memsetXX implementation is completely broken and fails in the first iteration, when i, j, and k are all zero, the failure is masked as zero is returned. Failing in the first iteration is perhaps the most likely failure, so this makes the tests pretty much useless. Avoid the situation by always setting a random unused bit in the result on failure.

Fixes 03270c13c5ff ("lib/string.c: add testcases for memset16/32/64")
---
 lib/test_string.c | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
lib/test_string.c +3 -3
index 0fcdb82dca86..98a787e7a1fd 100644
@@ -35,7 +35,7 @@ static __init int memset16_selftest(void)
 fail:
 	kfree(p);
 	if (i < 256)
-		return (i << 24) | (j << 16) | k;
+		return (i << 24) | (j << 16) | k | 0x8000;
 	return 0;
 }
 
@@ -71,7 +71,7 @@ static __init int memset32_selftest(void)
 fail:
 	kfree(p);
 	if (i < 256)
-		return (i << 24) | (j << 16) | k;
+		return (i << 24) | (j << 16) | k | 0x8000;
 	return 0;
 }
 
@@ -107,7 +107,7 @@ static __init int memset64_selftest(void)
 fail:
 	kfree(p);
 	if (i < 256)
-		return (i << 24) | (j << 16) | k;
+		return (i << 24) | (j << 16) | k | 0x8000;
 	return 0;
 }
 
Signature · 1 line
2.20.1
author maintainer
#32 lore raw

PATCH 33/47 coda: add error handling for fget

In reply to Sasha Levin

From: Zhouyang Jia <[email protected]>

[ Upstream commit 02551c23bcd85f0c68a8259c7b953d49d44f86af ]

When fget fails, the lack of error-handling code may cause unexpected results.

This patch adds error-handling code after calling fget.

---
 fs/coda/psdev.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
fs/coda/psdev.c +4 -1
index c5234c21b539..55824cba3245 100644
@@ -187,8 +187,11 @@ static ssize_t coda_psdev_write(struct file *file, const char __user *buf,
 	if (req->uc_opcode == CODA_OPEN_BY_FD) {
 		struct coda_open_by_fd_out *outp =
 			(struct coda_open_by_fd_out *)req->uc_data;
-		if (!outp->oh.result)
+		if (!outp->oh.result) {
 			outp->fh = fget(outp->fd);
+			if (!outp->fh)
+				return -EBADF;
+		}
 	}
 
         wake_up(&req->uc_sleep);
Signature · 1 line
2.20.1
author maintainer
#33 lore raw

PATCH 34/47 coda: fix build using bare-metal toolchain

In reply to Sasha Levin

From: Sam Protsenko <[email protected]>

[ Upstream commit b2a57e334086602be56b74958d9f29b955cd157f ]

The kernel is self-contained project and can be built with bare-metal toolchain. But bare-metal toolchain doesn't define __linux__. Because of this u_quad_t type is not defined when using bare-metal toolchain and codafs build fails. This patch fixes it by defining u_quad_t type unconditionally.

---
 include/linux/coda.h | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
include/linux/coda.h +1 -2
index d30209b9cef8..0ca0c83fdb1c 100644
@@ -58,8 +58,7 @@ Mellon the rights to redistribute these changes without encumbrance.
 #ifndef _CODA_HEADER_
 #define _CODA_HEADER_
 
-#if defined(__linux__)
 typedef unsigned long long u_quad_t;
-#endif
+
 #include <uapi/linux/coda.h>
 #endif 
Signature · 1 line
2.20.1
author maintainer
#34 lore raw

PATCH 35/47 uapi linux/coda_psdev.h: move upc_req definition from uapi to kernel side headers

In reply to Sasha Levin

From: Mikko Rapeli <[email protected]>

[ Upstream commit f90fb3c7e2c13ae829db2274b88b845a75038b8a ]

Only users of upc_req in kernel side fs/coda/psdev.c and fs/coda/upcall.c already include linux/coda_psdev.h.

Suggested by Jan Harkes <[email protected]> in
https://lore.kernel.org/lkml/[email protected]/

Fixes these include/uapi/linux/coda_psdev.h compilation errors in userspace:

linux/coda_psdev.h:12:19: error: field `uc_chain' has incomplete type

  struct list_head    uc_chain;
                   ^

linux/coda_psdev.h:13:2: error: unknown type name `caddr_t'
caddr_t uc_data;
^
linux/coda_psdev.h:14:2: error: unknown type name `u_short'
u_short uc_flags;
^
linux/coda_psdev.h:15:2: error: unknown type name `u_short'
u_short uc_inSize; /* Size is at most 5000 bytes */
^
linux/coda_psdev.h:16:2: error: unknown type name `u_short'
u_short uc_outSize;
^
linux/coda_psdev.h:17:2: error: unknown type name `u_short'
u_short uc_opcode; /* copied from data to save lookup */
^
linux/coda_psdev.h:19:2: error: unknown type name `wait_queue_head_t'
wait_queue_head_t uc_sleep; /* process' wait queue */
^

---
 include/linux/coda_psdev.h      | 11 +++++++++++
 include/uapi/linux/coda_psdev.h | 13 -------------
 2 files changed, 11 insertions(+), 13 deletions(-)
include/linux/coda_psdev.h +11 -0
index 15170954aa2b..57d2b2faf6a3 100644
@@ -19,6 +19,17 @@ struct venus_comm {
 	struct mutex	    vc_mutex;
 };
 
+/* messages between coda filesystem in kernel and Venus */
+struct upc_req {
+	struct list_head	uc_chain;
+	caddr_t			uc_data;
+	u_short			uc_flags;
+	u_short			uc_inSize;  /* Size is at most 5000 bytes */
+	u_short			uc_outSize;
+	u_short			uc_opcode;  /* copied from data to save lookup */
+	int			uc_unique;
+	wait_queue_head_t	uc_sleep;   /* process' wait queue */
+};
 
 static inline struct venus_comm *coda_vcp(struct super_block *sb)
 {
include/uapi/linux/coda_psdev.h +0 -13
index aa6623efd2dd..d50d51a57fe4 100644
@@ -7,19 +7,6 @@
 #define CODA_PSDEV_MAJOR 67
 #define MAX_CODADEVS  5	   /* how many do we allow */
 
-
-/* messages between coda filesystem in kernel and Venus */
-struct upc_req {
-	struct list_head    uc_chain;
-	caddr_t	            uc_data;
-	u_short	            uc_flags;
-	u_short             uc_inSize;  /* Size is at most 5000 bytes */
-	u_short	            uc_outSize;
-	u_short	            uc_opcode;  /* copied from data to save lookup */
-	int		    uc_unique;
-	wait_queue_head_t   uc_sleep;   /* process' wait queue */
-};
-
 #define CODA_REQ_ASYNC  0x1
 #define CODA_REQ_READ   0x2
 #define CODA_REQ_WRITE  0x4
Signature · 1 line
2.20.1
author maintainer
#35 lore raw

PATCH 36/47 drivers/rapidio/devices/rio_mport_cdev.c: NUL terminate some strings

In reply to Sasha Levin

From: Dan Carpenter <[email protected]>

[ Upstream commit 156e0b1a8112b76e351684ac948c59757037ac36 ]

The dev_info.name[] array has space for RIO_MAX_DEVNAME_SZ + 1 characters. But the problem here is that we don't ensure that the user put a NUL terminator on the end of the string. It could lead to an out of bounds read.

---
 drivers/rapidio/devices/rio_mport_cdev.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/rapidio/devices/rio_mport_cdev.c +2 -0
index cbe467ff1aba..fa0bbda4b3f2 100644
@@ -1688,6 +1688,7 @@ static int rio_mport_add_riodev(struct mport_cdev_priv *priv,
 
 	if (copy_from_user(&dev_info, arg, sizeof(dev_info)))
 		return -EFAULT;
+	dev_info.name[sizeof(dev_info.name) - 1] = '\0';
 
 	rmcd_debug(RDEV, "name:%s ct:0x%x did:0x%x hc:0x%x", dev_info.name,
 		   dev_info.comptag, dev_info.destid, dev_info.hopcount);
@@ -1819,6 +1820,7 @@ static int rio_mport_del_riodev(struct mport_cdev_priv *priv, void __user *arg)
 
 	if (copy_from_user(&dev_info, arg, sizeof(dev_info)))
 		return -EFAULT;
+	dev_info.name[sizeof(dev_info.name) - 1] = '\0';
 
 	mport = priv->md->mport;
 
Signature · 1 line
2.20.1
author maintainer
#36 lore raw

PATCH 37/47 drivers/pps/pps.c: clear offset flags in PPS_SETPARAMS ioctl

In reply to Sasha Levin

From: Miroslav Lichvar <[email protected]>

[ Upstream commit 5515e9a6273b8c02034466bcbd717ac9f53dab99 ]

The PPS assert/clear offset corrections are set by the PPS_SETPARAMS ioctl in the pps_ktime structs, which also contain flags. The flags are not initialized by applications (using the timepps.h header) and they are not used by the kernel for anything except returning them back in the PPS_GETPARAMS ioctl.

Set the flags to zero to make it clear they are unused and avoid leaking uninitialized data of the PPS_SETPARAMS caller to other applications that have a read access to the PPS device.

---
 drivers/pps/pps.c | 8 ++++++++
 1 file changed, 8 insertions(+)
drivers/pps/pps.c +8 -0
index 8febacb8fc54..0951564b6830 100644
@@ -166,6 +166,14 @@ static long pps_cdev_ioctl(struct file *file,
 			pps->params.mode |= PPS_CANWAIT;
 		pps->params.api_version = PPS_API_VERS;
 
+		/*
+		 * Clear unused fields of pps_kparams to avoid leaking
+		 * uninitialized data of the PPS_SETPARAMS caller via
+		 * PPS_GETPARAMS
+		 */
+		pps->params.assert_off_tu.flags = 0;
+		pps->params.clear_off_tu.flags = 0;
+
 		spin_unlock_irq(&pps->lock);
 
 		break;
Signature · 1 line
2.20.1
author maintainer
#37 lore raw

PATCH 38/47 ipc/mqueue.c: only perform resource calculation if user valid

In reply to Sasha Levin

From: Kees Cook <[email protected]>

[ Upstream commit a318f12ed8843cfac53198390c74a565c632f417 ]

Andreas Christoforou reported:

  UBSAN: Undefined behaviour in ipc/mqueue.c:414:49 signed integer overflow:
  9 * 2305843009213693951 cannot be represented in type 'long int'
  ...
  Call Trace:
    mqueue_evict_inode+0x8e7/0xa10 ipc/mqueue.c:414
    evict+0x472/0x8c0 fs/inode.c:558
    iput_final fs/inode.c:1547 [inline]
    iput+0x51d/0x8c0 fs/inode.c:1573
    mqueue_get_inode+0x8eb/0x1070 ipc/mqueue.c:320
    mqueue_create_attr+0x198/0x440 ipc/mqueue.c:459
    vfs_mkobj+0x39e/0x580 fs/namei.c:2892
    prepare_open ipc/mqueue.c:731 [inline]
    do_mq_open+0x6da/0x8e0 ipc/mqueue.c:771

Which could be triggered by:

        struct mq_attr attr = {
                .mq_flags = 0,
                .mq_maxmsg = 9,
                .mq_msgsize = 0x1fffffffffffffff,
                .mq_curmsgs = 0,
        };

        if (mq_open("/testing", 0x40, 3, &attr) == (mqd_t) -1)
                perror("mq_open");

mqueue_get_inode() was correctly rejecting the giant mq_msgsize, and preparing to return -EINVAL. During the cleanup, it calls mqueue_evict_inode() which performed resource usage tracking math for updating "user", before checking if there was a valid "user" at all (which would indicate that the calculations would be sane). Instead, delay this check to after seeing a valid "user".

The overflow was real, but the results went unused, so while the flaw is harmless, it's noisy for kernel fuzzers, so just fix it by moving the calculation under the non-NULL "user" where it actually gets used.

---
 ipc/mqueue.c | 19 ++++++++++---------
 1 file changed, 10 insertions(+), 9 deletions(-)
ipc/mqueue.c +10 -9
index bce7af1546d9..de4070d5472f 100644
@@ -389,7 +389,6 @@ static void mqueue_evict_inode(struct inode *inode)
 {
 	struct mqueue_inode_info *info;
 	struct user_struct *user;
-	unsigned long mq_bytes, mq_treesize;
 	struct ipc_namespace *ipc_ns;
 	struct msg_msg *msg, *nmsg;
 	LIST_HEAD(tmp_msg);
@@ -412,16 +411,18 @@ static void mqueue_evict_inode(struct inode *inode)
 		free_msg(msg);
 	}
 
-	/* Total amount of bytes accounted for the mqueue */
-	mq_treesize = info->attr.mq_maxmsg * sizeof(struct msg_msg) +
-		min_t(unsigned int, info->attr.mq_maxmsg, MQ_PRIO_MAX) *
-		sizeof(struct posix_msg_tree_node);
-
-	mq_bytes = mq_treesize + (info->attr.mq_maxmsg *
-				  info->attr.mq_msgsize);
-
 	user = info->user;
 	if (user) {
+		unsigned long mq_bytes, mq_treesize;
+
+		/* Total amount of bytes accounted for the mqueue */
+		mq_treesize = info->attr.mq_maxmsg * sizeof(struct msg_msg) +
+			min_t(unsigned int, info->attr.mq_maxmsg, MQ_PRIO_MAX) *
+			sizeof(struct posix_msg_tree_node);
+
+		mq_bytes = mq_treesize + (info->attr.mq_maxmsg *
+					  info->attr.mq_msgsize);
+
 		spin_lock(&mq_lock);
 		user->mq_bytes -= mq_bytes;
 		/*
Signature · 1 line
2.20.1
author maintainer
#38 lore raw

PATCH 39/47 floppy: fix div-by-zero in setup_format_params

In reply to Sasha Levin

From: Denis Efremov <[email protected]>

[ Upstream commit f3554aeb991214cbfafd17d55e2bfddb50282e32 ]

This fixes a divide by zero error in the setup_format_params function of the floppy driver.

Two consecutive ioctls can trigger the bug: The first one should set the drive geometry with such .sect and .rate values for the F_SECT_PER_TRACK to become zero. Next, the floppy format operation should be called.

A floppy disk is not required to be inserted. An unprivileged user could trigger the bug if the device is accessible.

The patch checks F_SECT_PER_TRACK for a non-zero value in the set_geometry function. The proper check should involve a reasonable upper limit for the .sect and .rate fields, but it could change the UAPI.

The patch also checks F_SECT_PER_TRACK in the setup_format_params, and cancels the formatting operation in case of zero.

The bug was found by syzkaller.

---
 drivers/block/floppy.c | 5 +++++
 1 file changed, 5 insertions(+)
drivers/block/floppy.c +5 -0
index a8de56f1936d..b1425b218606 100644
@@ -2119,6 +2119,9 @@ static void setup_format_params(int track)
 	raw_cmd->kernel_data = floppy_track_buffer;
 	raw_cmd->length = 4 * F_SECT_PER_TRACK;
 
+	if (!F_SECT_PER_TRACK)
+		return;
+
 	/* allow for about 30ms for data transport per track */
 	head_shift = (F_SECT_PER_TRACK + 5) / 6;
 
@@ -3243,6 +3246,8 @@ static int set_geometry(unsigned int cmd, struct floppy_struct *g,
 	/* sanity checking for parameters. */
 	if (g->sect <= 0 ||
 	    g->head <= 0 ||
+	    /* check for zero in F_SECT_PER_TRACK */
+	    (unsigned char)((g->sect << 2) >> FD_SIZECODE(g)) == 0 ||
 	    g->track <= 0 || g->track > UDP->tracks >> STRETCH(g) ||
 	    /* check if reserved bits are set */
 	    (g->stretch & ~(FD_STRETCH | FD_SWAPSIDES | FD_SECTBASEMASK)) != 0)
Signature · 1 line
2.20.1
author maintainer
#39 lore raw

PATCH 40/47 floppy: fix out-of-bounds read in copy_buffer

In reply to Sasha Levin

From: Denis Efremov <[email protected]>

[ Upstream commit da99466ac243f15fbba65bd261bfc75ffa1532b6 ]

This fixes a global out-of-bounds read access in the copy_buffer function of the floppy driver.

The FDDEFPRM ioctl allows one to set the geometry of a disk. The sect and head fields (unsigned int) of the floppy_drive structure are used to compute the max_sector (int) in the make_raw_rw_request function. It is possible to overflow the max_sector. Next, max_sector is passed to the copy_buffer function and used in one of the memcpy calls.

An unprivileged user could trigger the bug if the device is accessible, but requires a floppy disk to be inserted.

The patch adds the check for the .sect * .head multiplication for not overflowing in the set_geometry function.

The bug was found by syzkaller.

---
 drivers/block/floppy.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/block/floppy.c +4 -2
index b1425b218606..0d43e90eb252 100644
@@ -3244,8 +3244,10 @@ static int set_geometry(unsigned int cmd, struct floppy_struct *g,
 	int cnt;
 
 	/* sanity checking for parameters. */
-	if (g->sect <= 0 ||
-	    g->head <= 0 ||
+	if ((int)g->sect <= 0 ||
+	    (int)g->head <= 0 ||
+	    /* check for overflow in max_sector */
+	    (int)(g->sect * g->head) <= 0 ||
 	    /* check for zero in F_SECT_PER_TRACK */
 	    (unsigned char)((g->sect << 2) >> FD_SIZECODE(g)) == 0 ||
 	    g->track <= 0 || g->track > UDP->tracks >> STRETCH(g) ||
Signature · 1 line
2.20.1
author maintainer
#40 lore raw

PATCH 41/47 mlxsw: spectrum_dcb: Configure DSCP map as the last rule is removed

In reply to Sasha Levin

From: Petr Machata <[email protected]>

[ Upstream commit dedfde2fe1c4ccf27179fcb234e2112d065c39bb ]

Spectrum systems use DSCP rewrite map to update DSCP field in egressing packets to correspond to priority that the packet has. Whether rewriting will take place is determined at the point when the packet ingresses the
switch: if the port is in Trust L3 mode, packet priority is determined from the DSCP map at the port, and DSCP rewrite will happen. If the port is in Trust L2 mode, 802.1p is used for packet prioritization, and no DSCP rewrite will happen.

The driver determines the port trust mode based on whether any DSCP prioritization rules are in effect at given port. If there are any, trust level is L3, otherwise it's L2. When the last DSCP rule is removed, the port is switched to trust L2. Under that scenario, if DSCP of a packet should be rewritten, it should be rewritten to 0.

However, when switching to Trust L2, the driver neglects to also update the DSCP rewrite map. The last DSCP rule thus remains in effect, and packets egressing through this port, if they have the right priority, will have their DSCP set according to this rule.

Fix by first configuring the rewrite map, and only then switching to trust L2 and bailing out.

Fixes b2b1dab6884e ("mlxsw: spectrum: Support ieee_setapp, ieee_delapp")
---
 .../net/ethernet/mellanox/mlxsw/spectrum_dcb.c   | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/net/ethernet/mellanox/mlxsw/spectrum_dcb.c +8 -8
index b25048c6c761..21296fa7f7fb 100644
@@ -408,14 +408,6 @@ static int mlxsw_sp_port_dcb_app_update(struct mlxsw_sp_port *mlxsw_sp_port)
 	have_dscp = mlxsw_sp_port_dcb_app_prio_dscp_map(mlxsw_sp_port,
 							&prio_map);
 
-	if (!have_dscp) {
-		err = mlxsw_sp_port_dcb_toggle_trust(mlxsw_sp_port,
-					MLXSW_REG_QPTS_TRUST_STATE_PCP);
-		if (err)
-			netdev_err(mlxsw_sp_port->dev, "Couldn't switch to trust L2\n");
-		return err;
-	}
-
 	mlxsw_sp_port_dcb_app_dscp_prio_map(mlxsw_sp_port, default_prio,
 					    &dscp_map);
 	err = mlxsw_sp_port_dcb_app_update_qpdpm(mlxsw_sp_port,
@@ -432,6 +424,14 @@ static int mlxsw_sp_port_dcb_app_update(struct mlxsw_sp_port *mlxsw_sp_port)
 		return err;
 	}
 
+	if (!have_dscp) {
+		err = mlxsw_sp_port_dcb_toggle_trust(mlxsw_sp_port,
+					MLXSW_REG_QPTS_TRUST_STATE_PCP);
+		if (err)
+			netdev_err(mlxsw_sp_port->dev, "Couldn't switch to trust L2\n");
+		return err;
+	}
+
 	err = mlxsw_sp_port_dcb_toggle_trust(mlxsw_sp_port,
 					     MLXSW_REG_QPTS_TRUST_STATE_DSCP);
 	if (err) {
Signature · 1 line
2.20.1
author maintainer
#41 lore raw

PATCH 42/47 xen: let alloc_xenballooned_pages() fail if not enough memory free

In reply to Sasha Levin

From: Juergen Gross <[email protected]>

[ Upstream commit a1078e821b605813b63bf6bca414a85f804d5c66 ]

Instead of trying to allocate pages with GFP_USER in add_ballooned_pages() check the available free memory via si_mem_available(). GFP_USER is far less limiting memory exhaustion than the test via si_mem_available().

This will avoid dom0 running out of memory due to excessive foreign page mappings especially on ARM and on x86 in PVH mode, as those don't have a pre-ballooned area which can be used for foreign mappings.

As the normal ballooning suffers from the same problem don't balloon down more than si_mem_available() pages in one iteration. At the same time limit the default maximum number of retries.

This is part of XSA-300.

---
 drivers/xen/balloon.c | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)
drivers/xen/balloon.c +13 -3
index 7ab6caef599c..d4e8b717ce2b 100644
@@ -527,8 +527,15 @@ static void balloon_process(struct work_struct *work)
 				state = reserve_additional_memory();
 		}
 
-		if (credit < 0)
-			state = decrease_reservation(-credit, GFP_BALLOON);
+		if (credit < 0) {
+			long n_pages;
+
+			n_pages = min(-credit, si_mem_available());
+			state = decrease_reservation(n_pages, GFP_BALLOON);
+			if (state == BP_DONE && n_pages != -credit &&
+			    n_pages < totalreserve_pages)
+				state = BP_EAGAIN;
+		}
 
 		state = update_schedule(state);
 
@@ -567,6 +574,9 @@ static int add_ballooned_pages(int nr_pages)
 		}
 	}
 
+	if (si_mem_available() < nr_pages)
+		return -ENOMEM;
+
 	st = decrease_reservation(nr_pages, GFP_USER);
 	if (st != BP_DONE)
 		return -ENOMEM;
@@ -696,7 +706,7 @@ static int __init balloon_init(void)
 	balloon_stats.schedule_delay = 1;
 	balloon_stats.max_schedule_delay = 32;
 	balloon_stats.retry_count = 1;
-	balloon_stats.max_retry_count = RETRY_UNLIMITED;
+	balloon_stats.max_retry_count = 4;
 
 #ifdef CONFIG_XEN_BALLOON_MEMORY_HOTPLUG
 	set_online_page_callback(&xen_online_page);
Signature · 1 line
2.20.1
author maintainer
#42 lore raw

PATCH 43/47 xen/pv: Fix a boot up hang revealed by int3 self test

In reply to Sasha Levin

From: Zhenzhong Duan <[email protected]>

[ Upstream commit b23e5844dfe78a80ba672793187d3f52e4b528d7 ]

Commit 7457c0da024b ("x86/alternatives: Add int3_emulate_call() selftest") is used to ensure there is a gap setup in int3 exception stack which could be used for inserting call return address.

This gap is missed in XEN PV int3 exception entry path, then below panic triggered:

[    0.772876] general protection fault: 0000 [#1] SMP NOPTI
[    0.772886] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.2.0+ #11
[    0.772893] RIP: e030:int3_magic+0x0/0x7
[    0.772905] RSP: 3507:ffffffff82203e98 EFLAGS: 00000246
[    0.773334] Call Trace:
[    0.773334]  alternative_instructions+0x3d/0x12e
[    0.773334]  check_bugs+0x7c9/0x887
[    0.773334]  ? __get_locked_pte+0x178/0x1f0
[    0.773334]  start_kernel+0x4ff/0x535
[    0.773334]  ? set_init_arg+0x55/0x55
[    0.773334]  xen_start_kernel+0x571/0x57a

For 64bit PV guests, Xen's ABI enters the kernel with using SYSRET, with %rcx/%r11 on the stack. To convert back to "normal" looking exceptions, the xen thunks do 'xen_*: pop %rcx; pop %r11; jmp *'.

E.g. Extracting 'xen_pv_trap xenint3' we have:
xen_xenint3:
pop %rcx;
pop %r11;
jmp xenint3

As xenint3 and int3 entry code are same except xenint3 doesn't generate a gap, we can fix it by using int3 and drop useless xenint3.

Cc Boris Ostrovsky <[email protected]>
Cc Juergen Gross <[email protected]>
Cc Stefano Stabellini <[email protected]>
Cc Andy Lutomirski <[email protected]>
Cc Peter Zijlstra <[email protected]>
Cc Thomas Gleixner <[email protected]>
Cc Ingo Molnar <[email protected]>
Cc Borislav Petkov <[email protected]>
Cc Andrew Cooper <[email protected]>
---
 arch/x86/entry/entry_64.S    | 1 -
 arch/x86/include/asm/traps.h | 2 +-
 arch/x86/xen/enlighten_pv.c  | 2 +-
 arch/x86/xen/xen-asm_64.S    | 1 -
 4 files changed, 2 insertions(+), 4 deletions(-)
arch/x86/entry/entry_64.S +0 -1
index 206df099950e..e7572a209fbe 100644
@@ -1196,7 +1196,6 @@ idtentry stack_segment		do_stack_segment	has_error_code=1
 #ifdef CONFIG_XEN
 idtentry xennmi			do_nmi			has_error_code=0
 idtentry xendebug		do_debug		has_error_code=0
-idtentry xenint3		do_int3			has_error_code=0
 #endif
 
 idtentry general_protection	do_general_protection	has_error_code=1
arch/x86/include/asm/traps.h +1 -1
index afbc87206886..b771bb3d159b 100644
@@ -40,7 +40,7 @@ asmlinkage void simd_coprocessor_error(void);
 asmlinkage void xen_divide_error(void);
 asmlinkage void xen_xennmi(void);
 asmlinkage void xen_xendebug(void);
-asmlinkage void xen_xenint3(void);
+asmlinkage void xen_int3(void);
 asmlinkage void xen_overflow(void);
 asmlinkage void xen_bounds(void);
 asmlinkage void xen_invalid_op(void);
arch/x86/xen/enlighten_pv.c +1 -1
index 782f98b332f0..1730a26ff6ab 100644
@@ -597,12 +597,12 @@ struct trap_array_entry {
 
 static struct trap_array_entry trap_array[] = {
 	{ debug,                       xen_xendebug,                    true },
-	{ int3,                        xen_xenint3,                     true },
 	{ double_fault,                xen_double_fault,                true },
 #ifdef CONFIG_X86_MCE
 	{ machine_check,               xen_machine_check,               true },
 #endif
 	{ nmi,                         xen_xennmi,                      true },
+	{ int3,                        xen_int3,                        false },
 	{ overflow,                    xen_overflow,                    false },
 #ifdef CONFIG_IA32_EMULATION
 	{ entry_INT80_compat,          xen_entry_INT80_compat,          false },
arch/x86/xen/xen-asm_64.S +0 -1
index 417b339e5c8e..3a6feed76dfc 100644
@@ -30,7 +30,6 @@ xen_pv_trap divide_error
 xen_pv_trap debug
 xen_pv_trap xendebug
 xen_pv_trap int3
-xen_pv_trap xenint3
 xen_pv_trap xennmi
 xen_pv_trap overflow
 xen_pv_trap bounds
Signature · 1 line
2.20.1
author maintainer
#43 lore raw

PATCH 44/47 x86/kvm: Don't call kvm_spurious_fault() from .fixup

In reply to Sasha Levin

From: Josh Poimboeuf <[email protected]>

[ Upstream commit 3901336ed9887b075531bffaeef7742ba614058b ]

After making a change to improve objtool's sibling call detection, it started showing the following warning:

arch/x86/kvm/vmx/nested.o: warning: objtool: .fixup+0x15: sibling call from callable instruction with modified stack frame

The problem is the ____kvm_handle_fault_on_reboot() macro. It does a fake call by pushing a fake RIP and doing a jump. That tricks the unwinder into printing the function which triggered the exception, rather than the .fixup code.

Instead of the hack to make it look like the original function made the call, just change the macro so that the original function actually does make the call. This allows removal of the hack, and also makes objtool happy.

I triggered a vmx instruction exception and verified that the stack trace is still sane:

kernel BUG at arch/x86/kvm/x86.c:358!
invalid opcode: 0000 [#1] SMP PTI

  CPU: 28 PID: 4096 Comm: qemu-kvm Not tainted 5.2.0+ #16
  Hardware name: Lenovo THINKSYSTEM SD530 -[7X2106Z000]-/-[7X2106Z000]-, BIOS -[TEE113Z-1.00]- 07/17/2017
  RIP: 0010:kvm_spurious_fault+0x5/0x10
  Code: 00 00 00 00 00 8b 44 24 10 89 d2 45 89 c9 48 89 44 24 10 8b 44 24 08 48 89 44 24 08 e9 d4 40 22 00 0f 1f 40 00 0f 1f 44 00 00 <0f> 0b 66 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 41 55 49 89 fd 41
  RSP: 0018:ffffbf91c683bd00 EFLAGS: 00010246
  RAX: 000061f040000000 RBX: ffff9e159c77bba0 RCX: ffff9e15a5c87000
  RDX: 0000000665c87000 RSI: ffff9e15a5c87000 RDI: ffff9e159c77bba0
  RBP: 0000000000000000 R08: 0000000000000000 R09: ffff9e15a5c87000
  R10: 0000000000000000 R11: fffff8f2d99721c0 R12: ffff9e159c77bba0
  R13: ffffbf91c671d960 R14: ffff9e159c778000 R15: 0000000000000000
  FS:  00007fa341cbe700(0000) GS:ffff9e15b7400000(0000) knlGS:0000000000000000
  CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
  CR2: 00007fdd38356804 CR3: 00000006759de003 CR4: 00000000007606e0
  DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
  DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
  PKRU: 55555554
  Call Trace:
   loaded_vmcs_init+0x4f/0xe0
   alloc_loaded_vmcs+0x38/0xd0
   vmx_create_vcpu+0xf7/0x600
   kvm_vm_ioctl+0x5e9/0x980
   ? __switch_to_asm+0x40/0x70
   ? __switch_to_asm+0x34/0x70
   ? __switch_to_asm+0x40/0x70
   ? __switch_to_asm+0x34/0x70
   ? free_one_page+0x13f/0x4e0
   do_vfs_ioctl+0xa4/0x630
   ksys_ioctl+0x60/0x90
   __x64_sys_ioctl+0x16/0x20
   do_syscall_64+0x55/0x1c0
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  RIP: 0033:0x7fa349b1ee5b
---
 arch/x86/include/asm/kvm_host.h | 34 ++++++++++++++++++---------------
 1 file changed, 19 insertions(+), 15 deletions(-)
arch/x86/include/asm/kvm_host.h +19 -15
index 7014dba23d20..2877e1fbadd8 100644
@@ -1427,25 +1427,29 @@ enum {
 #define kvm_arch_vcpu_memslots_id(vcpu) ((vcpu)->arch.hflags & HF_SMM_MASK ? 1 : 0)
 #define kvm_memslots_for_spte_role(kvm, role) __kvm_memslots(kvm, (role).smm)
 
+asmlinkage void __noreturn kvm_spurious_fault(void);
+
 /*
  * Hardware virtualization extension instructions may fault if a
  * reboot turns off virtualization while processes are running.
- * Trap the fault and ignore the instruction if that happens.
+ * Usually after catching the fault we just panic; during reboot
+ * instead the instruction is ignored.
  */
-asmlinkage void kvm_spurious_fault(void);
-
-#define ____kvm_handle_fault_on_reboot(insn, cleanup_insn)	\
-	"666: " insn "\n\t" \
-	"668: \n\t"                           \
-	".pushsection .fixup, \"ax\" \n" \
-	"667: \n\t" \
-	cleanup_insn "\n\t"		      \
-	"cmpb $0, kvm_rebooting \n\t"	      \
-	"jne 668b \n\t"      		      \
-	__ASM_SIZE(push) " $666b \n\t"	      \
-	"jmp kvm_spurious_fault \n\t"	      \
-	".popsection \n\t" \
-	_ASM_EXTABLE(666b, 667b)
+#define ____kvm_handle_fault_on_reboot(insn, cleanup_insn)		\
+	"666: \n\t"							\
+	insn "\n\t"							\
+	"jmp	668f \n\t"						\
+	"667: \n\t"							\
+	"call	kvm_spurious_fault \n\t"				\
+	"668: \n\t"							\
+	".pushsection .fixup, \"ax\" \n\t"				\
+	"700: \n\t"							\
+	cleanup_insn "\n\t"						\
+	"cmpb	$0, kvm_rebooting\n\t"					\
+	"je	667b \n\t"						\
+	"jmp	668b \n\t"						\
+	".popsection \n\t"						\
+	_ASM_EXTABLE(666b, 700b)
 
 #define __kvm_handle_fault_on_reboot(insn)		\
 	____kvm_handle_fault_on_reboot(insn, "")
Signature · 1 line
2.20.1
author maintainer
#44 lore raw

PATCH 45/47 x86/paravirt: Fix callee-saved function ELF sizes

In reply to Sasha Levin

From: Josh Poimboeuf <[email protected]>

[ Upstream commit 083db6764821996526970e42d09c1ab2f4155dd4 ]

The __raw_callee_save_*() functions have an ELF symbol size of zero, which confuses objtool and other tools.

Fixes a bunch of warnings like the following:

arch/x86/xen/mmu_pv.o: warning: objtool: __raw_callee_save_xen_pte_val() is missing an ELF size annotation arch/x86/xen/mmu_pv.o: warning: objtool: __raw_callee_save_xen_pgd_val() is missing an ELF size annotation arch/x86/xen/mmu_pv.o: warning: objtool: __raw_callee_save_xen_make_pte() is missing an ELF size annotation arch/x86/xen/mmu_pv.o: warning: objtool: __raw_callee_save_xen_make_pgd() is missing an ELF size annotation

---
 arch/x86/include/asm/paravirt.h | 1 +
 arch/x86/kernel/kvm.c           | 1 +
 2 files changed, 2 insertions(+)
arch/x86/include/asm/paravirt.h +1 -0
index e375d4266b53..a04677038872 100644
@@ -768,6 +768,7 @@ static __always_inline bool pv_vcpu_is_preempted(long cpu)
 	    PV_RESTORE_ALL_CALLER_REGS					\
 	    FRAME_END							\
 	    "ret;"							\
+	    ".size " PV_THUNK_NAME(func) ", .-" PV_THUNK_NAME(func) ";"	\
 	    ".popsection")
 
 /* Get a reference to a callee-save function */
arch/x86/kernel/kvm.c +1 -0
index 7f89d609095a..cee45d46e67d 100644
@@ -830,6 +830,7 @@ asm(
 "cmpb	$0, " __stringify(KVM_STEAL_TIME_preempted) "+steal_time(%rax);"
 "setne	%al;"
 "ret;"
+".size __raw_callee_save___kvm_vcpu_is_preempted, .-__raw_callee_save___kvm_vcpu_is_preempted;"
 ".popsection");
 
 #endif
Signature · 1 line
2.20.1
author maintainer
#45 lore raw

PATCH 46/47 x86, boot: Remove multiple copy of static function sanitize_boot_params()

In reply to Sasha Levin

From: Zhenzhong Duan <[email protected]>

[ Upstream commit 8c5477e8046ca139bac250386c08453da37ec1ae ]

Kernel build warns:
'sanitize_boot_params' defined but not used [-Wunused-function]

at below files:
arch/x86/boot/compressed/cmdline.c
arch/x86/boot/compressed/error.c
arch/x86/boot/compressed/early_serial_console.c
arch/x86/boot/compressed/acpi.c

That's becausethey each include misc.h which includes a definition of sanitize_boot_params() via bootparam_utils.h.

Remove the inclusion from misc.h and have the c file including bootparam_utils.h directly.

---
 arch/x86/boot/compressed/misc.c | 1 +
 arch/x86/boot/compressed/misc.h | 1 -
 2 files changed, 1 insertion(+), 1 deletion(-)
arch/x86/boot/compressed/misc.c +1 -0
index 8dd1d5ccae58..0387d7a96c84 100644
@@ -17,6 +17,7 @@
 #include "pgtable.h"
 #include "../string.h"
 #include "../voffset.h"
+#include <asm/bootparam_utils.h>
 
 /*
  * WARNING!!
arch/x86/boot/compressed/misc.h +0 -1
index a423bdb42686..47fd18db6b3b 100644
@@ -22,7 +22,6 @@
 #include <asm/page.h>
 #include <asm/boot.h>
 #include <asm/bootparam.h>
-#include <asm/bootparam_utils.h>
 
 #define BOOT_BOOT_H
 #include "../ctype.h"
Signature · 1 line
2.20.1
author maintainer
#46 lore raw

PATCH 47/47 drm/nouveau: fix memory leak in nouveau_conn_reset()

In reply to Sasha Levin

From: Yongxin Liu <[email protected]>

[ Upstream commit 09b90e2fe35faeace2488234e2a7728f2ea8ba26 ]

In nouveau_conn_reset(), if connector->state is true, __drm_atomic_helper_connector_destroy_state() will be called, but the memory pointed by asyc isn't freed. Memory leak happens in the following function __drm_atomic_helper_connector_reset(), where newly allocated asyc->state will be assigned to connector->state.

So using nouveau_conn_atomic_destroy_state() instead of __drm_atomic_helper_connector_destroy_state to free the "old" asyc.

Here the is the log showing memory leak.

unreferenced object 0xffff8c5480483c80 (size 192):
comm "kworker/0:2", pid 188, jiffies 4294695279 (age 53.179s) hex dump (first 32 bytes):

    00 f0 ba 7b 54 8c ff ff 00 00 00 00 00 00 00 00  ...{T...........
    00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00  ................
  backtrace:
    [<000000005005c0d0>] kmem_cache_alloc_trace+0x195/0x2c0
    [<00000000a122baed>] nouveau_conn_reset+0x25/0xc0 [nouveau]
    [<000000004fd189a2>] nouveau_connector_create+0x3a7/0x610 [nouveau]
    [<00000000c73343a8>] nv50_display_create+0x343/0x980 [nouveau]
    [<000000002e2b03c3>] nouveau_display_create+0x51f/0x660 [nouveau]
    [<00000000c924699b>] nouveau_drm_device_init+0x182/0x7f0 [nouveau]
    [<00000000cc029436>] nouveau_drm_probe+0x20c/0x2c0 [nouveau]
    [<000000007e961c3e>] local_pci_probe+0x47/0xa0
    [<00000000da14d569>] work_for_cpu_fn+0x1a/0x30
    [<0000000028da4805>] process_one_work+0x27c/0x660
    [<000000001d415b04>] worker_thread+0x22b/0x3f0
    [<0000000003b69f1f>] kthread+0x12f/0x150
    [<00000000c94c29b7>] ret_from_fork+0x3a/0x50
---
 drivers/gpu/drm/nouveau/nouveau_connector.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/nouveau/nouveau_connector.c +1 -1
index 247f72cc4d10..fb0094fc5583 100644
@@ -251,7 +251,7 @@ nouveau_conn_reset(struct drm_connector *connector)
 		return;
 
 	if (connector->state)
-		__drm_atomic_helper_connector_destroy_state(connector->state);
+		nouveau_conn_atomic_destroy_state(connector, connector->state);
 	__drm_atomic_helper_connector_reset(connector, &asyc->state);
 	asyc->dither.mode = DITHERING_MODE_AUTO;
 	asyc->dither.depth = DITHERING_DEPTH_AUTO;
Signature · 6 lines
2.20.1

_______________________________________________
dri-devel mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/dri-devel