Skip to contentSkip to conversation

drm/tilcdc: fix leak & null ref in panel_connector_get_modes

PATCH In review Sasha Levin 3 participants 64 messages Aug 10, 2020 – Aug 15, 2020

Subsystems (22)
lore
Share & tools
lore

Sasha Levin

author maintainer
#0 lore raw

PATCH 1/60 drm/tilcdc: fix leak & null ref in panel_connector_get_modes

From: Tomi Valkeinen <[email protected]>

[ Upstream commit 3f9c1c872cc97875ddc8d63bc9fe6ee13652b933 ]

If videomode_from_timings() returns true, the mode allocated with drm_mode_create will be leaked.

Also, the return value of drm_mode_create() is never checked, and thus could cause NULL deref.

Fix these two issues.

---
 drivers/gpu/drm/tilcdc/tilcdc_panel.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/gpu/drm/tilcdc/tilcdc_panel.c +5 -1
index 5584e656b8575..8c4fd1aa4c2db 100644
@@ -143,12 +143,16 @@ static int panel_connector_get_modes(struct drm_connector *connector)
 	int i;
 
 	for (i = 0; i < timings->num_timings; i++) {
-		struct drm_display_mode *mode = drm_mode_create(dev);
+		struct drm_display_mode *mode;
 		struct videomode vm;
 
 		if (videomode_from_timings(timings, &vm, i))
 			break;
 
+		mode = drm_mode_create(dev);
+		if (!mode)
+			break;
+
 		drm_display_mode_from_videomode(&vm, mode);
 
 		mode->type = DRM_MODE_TYPE_DRIVER;
Signature · 6 lines
2.25.1

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

PATCH 2/60 soc: qcom: rpmh-rsc: Set suppress_bind_attrs flag

From: Maulik Shah <[email protected]>

[ Upstream commit 1a53ce9ab4faeb841b33d62d23283dc76c0e7c5a ]

rpmh-rsc driver is fairly core to system and should not be removable once its probed. However it allows to unbind driver from sysfs using below command which results into a crash on sc7180.

echo 18200000.rsc > /sys/bus/platform/drivers/rpmh/unbind

Lets prevent unbind at runtime by setting suppress_bind_attrs flag.

---
 drivers/soc/qcom/rpmh-rsc.c | 1 +
 1 file changed, 1 insertion(+)
drivers/soc/qcom/rpmh-rsc.c +1 -0
index 3d2104286ee91..a9ccdf2e43b78 100644
@@ -715,6 +715,7 @@ static struct platform_driver rpmh_driver = {
 	.driver = {
 		  .name = "rpmh",
 		  .of_match_table = rpmh_drv_match,
+		  .suppress_bind_attrs = true,
 	},
 };
 
Signature · 1 line
2.25.1
author maintainer
#2 lore raw

PATCH 3/60 net: mscc: ocelot: fix encoding destination ports into multicast IPv4 address

In reply to Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 0897ecf7532577bda3dbcb043ce046a96948889d ]

The ocelot hardware designers have made some hacks to support multicast IPv4 and IPv6 addresses. Normally, the MAC table matches on MAC addresses and the destination ports are selected through the DEST_IDX field of the respective MAC table entry. The DEST_IDX points to a Port Group ID (PGID) which contains the bit mask of ports that frames should be forwarded to. But there aren't a lot of PGIDs (only 80 or so) and there are clearly many more IP multicast addresses than that, so it doesn't scale to use this PGID mechanism, so something else was done. Since the first portion of the MAC address is known, the hack they did was to use a single PGID for _flooding_ unknown IPv4 multicast (PGID_MCIPV4 == 62), but for known IP multicast, embed the destination ports into the first 3 bytes of the MAC address recorded in the MAC table.

The VSC7514 datasheet explains it like this:

3.9.1.5 IPv4 Multicast Entries

MAC table entries with the ENTRY_TYPE = 2 settings are interpreted as IPv4 multicast entries.
IPv4 multicasts entries match IPv4 frames, which are classified to the specified VID, and which have DMAC = 0x01005Exxxxxx, where xxxxxx is the lower 24 bits of the MAC address in the entry. Instead of a lookup in the destination mask table (PGID), the destination set is programmed as part of the entry MAC address. This is shown in the following table.

Table 78: IPv4 Multicast Destination Mask

Destination Ports Record Bit Field
---------------------------------------------
Ports 10-0 MAC[34-24]

Example: All IPv4 multicast frames in VLAN 12 with MAC 01005E112233 are to be forwarded to ports 3, 8, and 9. This is done by inserting the following entry in the MAC table entry:
VALID = 1
VID = 12
MAC = 0x000308112233
ENTRY_TYPE = 2
DEST_IDX = 0

But this procedure is not at all what's going on in the driver. In fact, the code that embeds the ports into the MAC address looks like it hasn't actually been tested. This patch applies the procedure described in the datasheet.

Since there are many other fixes to be made around multicast forwarding until it works properly, there is no real reason for this patch to be backported to stable trees, or considered a real fix of something that should have worked.

---
 drivers/net/ethernet/mscc/ocelot.c | 16 ++++++++--------
 1 file changed, 8 insertions(+), 8 deletions(-)
drivers/net/ethernet/mscc/ocelot.c +8 -8
index efb3965a3e42b..b687e1caa0e17 100644
@@ -1599,14 +1599,14 @@ static int ocelot_port_obj_add_mdb(struct net_device *dev,
 	addr[0] = 0;
 
 	if (!new) {
-		addr[2] = mc->ports << 0;
-		addr[1] = mc->ports << 8;
+		addr[1] = mc->ports >> 8;
+		addr[2] = mc->ports & 0xff;
 		ocelot_mact_forget(ocelot, addr, vid);
 	}
 
 	mc->ports |= BIT(port);
-	addr[2] = mc->ports << 0;
-	addr[1] = mc->ports << 8;
+	addr[1] = mc->ports >> 8;
+	addr[2] = mc->ports & 0xff;
 
 	return ocelot_mact_learn(ocelot, 0, addr, vid, ENTRYTYPE_MACv4);
 }
@@ -1630,9 +1630,9 @@ static int ocelot_port_obj_del_mdb(struct net_device *dev,
 		return -ENOENT;
 
 	memcpy(addr, mc->addr, ETH_ALEN);
-	addr[2] = mc->ports << 0;
-	addr[1] = mc->ports << 8;
 	addr[0] = 0;
+	addr[1] = mc->ports >> 8;
+	addr[2] = mc->ports & 0xff;
 	ocelot_mact_forget(ocelot, addr, vid);
 
 	mc->ports &= ~BIT(port);
@@ -1642,8 +1642,8 @@ static int ocelot_port_obj_del_mdb(struct net_device *dev,
 		return 0;
 	}
 
-	addr[2] = mc->ports << 0;
-	addr[1] = mc->ports << 8;
+	addr[1] = mc->ports >> 8;
+	addr[2] = mc->ports & 0xff;
 
 	return ocelot_mact_learn(ocelot, 0, addr, vid, ENTRYTYPE_MACv4);
 }
Signature · 1 line
2.25.1
#3 lore raw

Re: [PATCH AUTOSEL 5.7 03/60] net: mscc: ocelot: fix encoding destination ports into multicast IPv4 address

Hi Sasha,

Quoted 56 lines · Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 0897ecf7532577bda3dbcb043ce046a96948889d ]

The ocelot hardware designers have made some hacks to support multicast IPv4 and IPv6 addresses. Normally, the MAC table matches on MAC addresses and the destination ports are selected through the DEST_IDX field of the respective MAC table entry. The DEST_IDX points to a Port Group ID (PGID) which contains the bit mask of ports that frames should be forwarded to. But there aren't a lot of PGIDs (only 80 or so) and there are clearly many more IP multicast addresses than that, so it doesn't scale to use this PGID mechanism, so something else was done. Since the first portion of the MAC address is known, the hack they did was to use a single PGID for _flooding_ unknown IPv4 multicast (PGID_MCIPV4 == 62), but for known IP multicast, embed the destination ports into the first 3 bytes of the MAC address recorded in the MAC table.

The VSC7514 datasheet explains it like this:

3.9.1.5 IPv4 Multicast Entries

MAC table entries with the ENTRY_TYPE = 2 settings are interpreted as IPv4 multicast entries.
IPv4 multicasts entries match IPv4 frames, which are classified to the specified VID, and which have DMAC = 0x01005Exxxxxx, where xxxxxx is the lower 24 bits of the MAC address in the entry. Instead of a lookup in the destination mask table (PGID), the destination set is programmed as part of the entry MAC address. This is shown in the following table.

Table 78: IPv4 Multicast Destination Mask

Destination Ports Record Bit Field
---------------------------------------------
Ports 10-0 MAC[34-24]

Example: All IPv4 multicast frames in VLAN 12 with MAC 01005E112233 are to be forwarded to ports 3, 8, and 9. This is done by inserting the following entry in the MAC table entry:
VALID = 1
VID = 12
MAC = 0x000308112233
ENTRY_TYPE = 2
DEST_IDX = 0

But this procedure is not at all what's going on in the driver. In fact, the code that embeds the ports into the MAC address looks like it hasn't actually been tested. This patch applies the procedure described in the datasheet.

Since there are many other fixes to be made around multicast forwarding until it works properly, there is no real reason for this patch to be backported to stable trees, or considered a real fix of something that should have worked.

Quoted 4 lines

---

Could you please drop this patch from the 'stable' queues for 5.7 and 5.8? I haven't tested it on older kernels and without the other patches sent in that series. I would like to avoid unexpected regressions if possible.

Thanks,
-Vladimir

author maintainer
#4 lore raw

Re: [PATCH AUTOSEL 5.7 03/60] net: mscc: ocelot: fix encoding destination ports into multicast IPv4 address

Quoted 64 lines · Vladimir Oltean

Hi Sasha,

Quoted 56 lines · Sasha Levin

From: Vladimir Oltean <[email protected]>

[ Upstream commit 0897ecf7532577bda3dbcb043ce046a96948889d ]

The ocelot hardware designers have made some hacks to support multicast IPv4 and IPv6 addresses. Normally, the MAC table matches on MAC addresses and the destination ports are selected through the DEST_IDX field of the respective MAC table entry. The DEST_IDX points to a Port Group ID (PGID) which contains the bit mask of ports that frames should be forwarded to. But there aren't a lot of PGIDs (only 80 or so) and there are clearly many more IP multicast addresses than that, so it doesn't scale to use this PGID mechanism, so something else was done. Since the first portion of the MAC address is known, the hack they did was to use a single PGID for _flooding_ unknown IPv4 multicast (PGID_MCIPV4 == 62), but for known IP multicast, embed the destination ports into the first 3 bytes of the MAC address recorded in the MAC table.

The VSC7514 datasheet explains it like this:

3.9.1.5 IPv4 Multicast Entries

MAC table entries with the ENTRY_TYPE = 2 settings are interpreted as IPv4 multicast entries.
IPv4 multicasts entries match IPv4 frames, which are classified to the specified VID, and which have DMAC = 0x01005Exxxxxx, where xxxxxx is the lower 24 bits of the MAC address in the entry. Instead of a lookup in the destination mask table (PGID), the destination set is programmed as part of the entry MAC address. This is shown in the following table.

Table 78: IPv4 Multicast Destination Mask

Destination Ports Record Bit Field
---------------------------------------------
Ports 10-0 MAC[34-24]

Example: All IPv4 multicast frames in VLAN 12 with MAC 01005E112233 are to be forwarded to ports 3, 8, and 9. This is done by inserting the following entry in the MAC table entry:
VALID = 1
VID = 12
MAC = 0x000308112233
ENTRY_TYPE = 2
DEST_IDX = 0

But this procedure is not at all what's going on in the driver. In fact, the code that embeds the ports into the MAC address looks like it hasn't actually been tested. This patch applies the procedure described in the datasheet.

Since there are many other fixes to be made around multicast forwarding until it works properly, there is no real reason for this patch to be backported to stable trees, or considered a real fix of something that should have worked.

Quoted 4 lines

---

Quoted 4 lines

Could you please drop this patch from the 'stable' queues for 5.7 and 5.8? I haven't tested it on older kernels and without the other patches sent in that series. I would like to avoid unexpected regressions if possible.

Will do, thanks!

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

PATCH 4/60 ARM: exynos: clear L310_AUX_CTRL_FULL_LINE_ZERO in default l2c_aux_val

In reply to Sasha Levin

From: Guillaume Tucker <[email protected]>

[ Upstream commit 5b17a04addc29201dc142c8d2c077eb7745d2e35 ]

This "alert" error message can be seen on exynos4412-odroidx2:

L2C: platform modifies aux control register: 0x02070000 -> 0x3e470001
L2C: platform provided aux values permit register corruption.

Followed by this plain error message:

L2C-310: enabling full line of zeros but not enabled in Cortex-A9

To fix it, don't set the L310_AUX_CTRL_FULL_LINE_ZERO flag (bit 0) in the default value of l2c_aux_val. It may instead be enabled when applicable by the logic in l2c310_enable() if the attribute "arm,full-line-zero-disable" was set in the device tree.

The initial commit that introduced this default value was in v2.6.38 commit 1cf0eb799759 ("ARM: S5PV310: Add L2 cache init function in cpu.c").

However, the code to set the L310_AUX_CTRL_FULL_LINE_ZERO flag and manage that feature was added much later and the default value was not updated then. So this seems to have been a subtle oversight especially since enabling it only in the cache and not in the A9 core doesn't actually prevent the platform from running. According to the TRM, the opposite would be a real issue, if the feature was enabled in the A9 core but not in the cache controller.

---
 arch/arm/mach-exynos/exynos.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/mach-exynos/exynos.c +1 -1
index 7a8d1555db404..36c37444485a8 100644
@@ -193,7 +193,7 @@ static void __init exynos_dt_fixup(void)
 }
 
 DT_MACHINE_START(EXYNOS_DT, "Samsung Exynos (Flattened Device Tree)")
-	.l2c_aux_val	= 0x3c400001,
+	.l2c_aux_val	= 0x3c400000,
 	.l2c_aux_mask	= 0xc20fffff,
 	.smp		= smp_ops(exynos_smp_ops),
 	.map_io		= exynos_init_io,
Signature · 1 line
2.25.1
author maintainer
#6 lore raw

PATCH 5/60 Bluetooth: add a mutex lock to avoid UAF in do_enale_set

In reply to Sasha Levin

From: Lihong Kou <[email protected]>

[ Upstream commit f9c70bdc279b191da8d60777c627702c06e4a37d ]

In the case we set or free the global value listen_chan in different threads, we can encounter the UAF problems because the method is not protected by any lock, add one to avoid this bug.

BUG: KASAN: use-after-free in l2cap_chan_close+0x48/0x990
net/bluetooth/l2cap_core.c:730
Read of size 8 at addr ffff888096950000 by task kworker/1:102/2868

CPU: 1 PID: 2868 Comm: kworker/1:102 Not tainted 5.5.0-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine,
BIOS Google 01/01/2011
Workqueue: events do_enable_set
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x1fb/0x318 lib/dump_stack.c:118
 print_address_description+0x74/0x5c0 mm/kasan/report.c:374
 __kasan_report+0x149/0x1c0 mm/kasan/report.c:506
 kasan_report+0x26/0x50 mm/kasan/common.c:641
 __asan_report_load8_noabort+0x14/0x20 mm/kasan/generic_report.c:135
 l2cap_chan_close+0x48/0x990 net/bluetooth/l2cap_core.c:730
 do_enable_set+0x660/0x900 net/bluetooth/6lowpan.c:1074
 process_one_work+0x7f5/0x10f0 kernel/workqueue.c:2264
 worker_thread+0xbbc/0x1630 kernel/workqueue.c:2410
 kthread+0x332/0x350 kernel/kthread.c:255
 ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352

Allocated by task 2870:
 save_stack mm/kasan/common.c:72 [inline]
 set_track mm/kasan/common.c:80 [inline]
 __kasan_kmalloc+0x118/0x1c0 mm/kasan/common.c:515
 kasan_kmalloc+0x9/0x10 mm/kasan/common.c:529
 kmem_cache_alloc_trace+0x221/0x2f0 mm/slab.c:3551
 kmalloc include/linux/slab.h:555 [inline]
 kzalloc include/linux/slab.h:669 [inline]
 l2cap_chan_create+0x50/0x320 net/bluetooth/l2cap_core.c:446
 chan_create net/bluetooth/6lowpan.c:640 [inline]
 bt_6lowpan_listen net/bluetooth/6lowpan.c:959 [inline]
 do_enable_set+0x6a4/0x900 net/bluetooth/6lowpan.c:1078
 process_one_work+0x7f5/0x10f0 kernel/workqueue.c:2264
 worker_thread+0xbbc/0x1630 kernel/workqueue.c:2410
 kthread+0x332/0x350 kernel/kthread.c:255
 ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352

Freed by task 2870:
 save_stack mm/kasan/common.c:72 [inline]
 set_track mm/kasan/common.c:80 [inline]
 kasan_set_free_info mm/kasan/common.c:337 [inline]
 __kasan_slab_free+0x12e/0x1e0 mm/kasan/common.c:476
 kasan_slab_free+0xe/0x10 mm/kasan/common.c:485
 __cache_free mm/slab.c:3426 [inline]
 kfree+0x10d/0x220 mm/slab.c:3757
 l2cap_chan_destroy net/bluetooth/l2cap_core.c:484 [inline]
 kref_put include/linux/kref.h:65 [inline]
 l2cap_chan_put+0x170/0x190 net/bluetooth/l2cap_core.c:498
 do_enable_set+0x66c/0x900 net/bluetooth/6lowpan.c:1075
 process_one_work+0x7f5/0x10f0 kernel/workqueue.c:2264
 worker_thread+0xbbc/0x1630 kernel/workqueue.c:2410
 kthread+0x332/0x350 kernel/kthread.c:255
 ret_from_fork+0x24/0x30 arch/x86/entry/entry_64.S:352

The buggy address belongs to the object at ffff888096950000
 which belongs to the cache kmalloc-2k of size 2048
The buggy address is located 0 bytes inside of
 2048-byte region [ffff888096950000, ffff888096950800)
The buggy address belongs to the page:
page:ffffea00025a5400 refcount:1 mapcount:0 mapping:ffff8880aa400e00 index:0x0
flags: 0xfffe0000000200(slab)
raw: 00fffe0000000200 ffffea00027d1548 ffffea0002397808 ffff8880aa400e00
raw: 0000000000000000 ffff888096950000 0000000100000001 0000000000000000
page dumped because: kasan: bad access detected

Memory state around the buggy address:
 ffff88809694ff00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
 ffff88809694ff80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
>ffff888096950000: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
                   ^
 ffff888096950080: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
 ffff888096950100: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
==================================================================
---
 net/bluetooth/6lowpan.c | 5 +++++
 1 file changed, 5 insertions(+)
net/bluetooth/6lowpan.c +5 -0
index 4febc82a7c761..52fb6d6d6d585 100644
@@ -50,6 +50,7 @@ static bool enable_6lowpan;
 /* We are listening incoming connections via this channel
  */
 static struct l2cap_chan *listen_chan;
+static DEFINE_MUTEX(set_lock);
 
 struct lowpan_peer {
 	struct list_head list;
@@ -1070,12 +1071,14 @@ static void do_enable_set(struct work_struct *work)
 
 	enable_6lowpan = set_enable->flag;
 
+	mutex_lock(&set_lock);
 	if (listen_chan) {
 		l2cap_chan_close(listen_chan, 0);
 		l2cap_chan_put(listen_chan);
 	}
 
 	listen_chan = bt_6lowpan_listen();
+	mutex_unlock(&set_lock);
 
 	kfree(set_enable);
 }
@@ -1127,11 +1130,13 @@ static ssize_t lowpan_control_write(struct file *fp,
 		if (ret == -EINVAL)
 			return ret;
 
+		mutex_lock(&set_lock);
 		if (listen_chan) {
 			l2cap_chan_close(listen_chan, 0);
 			l2cap_chan_put(listen_chan);
 			listen_chan = NULL;
 		}
+		mutex_unlock(&set_lock);
 
 		if (conn) {
 			struct lowpan_peer *peer;
Signature · 1 line
2.25.1
author maintainer
#7 lore raw

PATCH 6/60 loop: be paranoid on exit and prevent new additions / removals

In reply to Sasha Levin

From: Luis Chamberlain <[email protected]>

[ Upstream commit 200f93377220504c5e56754823e7adfea6037f1a ]

Be pedantic on removal as well and hold the mutex. This should prevent uses of addition while we exit.

---
 drivers/block/loop.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/block/loop.c +4 -0
index 418bb4621255a..6b36fc2f4edc7 100644
@@ -2333,6 +2333,8 @@ static void __exit loop_exit(void)
 
 	range = max_loop ? max_loop << part_shift : 1UL << MINORBITS;
 
+	mutex_lock(&loop_ctl_mutex);
+
 	idr_for_each(&loop_index_idr, &loop_exit_cb, NULL);
 	idr_destroy(&loop_index_idr);
 
@@ -2340,6 +2342,8 @@ static void __exit loop_exit(void)
 	unregister_blkdev(LOOP_MAJOR, "loop");
 
 	misc_deregister(&loop_misc);
+
+	mutex_unlock(&loop_ctl_mutex);
 }
 
 module_init(loop_init);
Signature · 1 line
2.25.1
author maintainer
#8 lore raw

PATCH 7/60 io_uring: fix req->work corruption

In reply to Sasha Levin

From: Pavel Begunkov <[email protected]>

[ Upstream commit 8ef77766ba8694968ed4ba24311b4bacee14f235 ]

req->work and req->task_work are in a union, so io_req_task_queue() screws everything that was in work. De-union them for now.

[  704.367253] BUG: unable to handle page fault for address:
	ffffffffaf7330d0
[  704.367256] #PF: supervisor write access in kernel mode
[  704.367256] #PF: error_code(0x0003) - permissions violation
[  704.367261] CPU: 6 PID: 1654 Comm: io_wqe_worker-0 Tainted: G
I       5.8.0-rc2-00038-ge28d0bdc4863-dirty #498
[  704.367265] RIP: 0010:_raw_spin_lock+0x1e/0x36
...
[  704.367276]  __alloc_fd+0x35/0x150
[  704.367279]  __get_unused_fd_flags+0x25/0x30
[  704.367280]  io_openat2+0xcb/0x1b0
[  704.367283]  io_issue_sqe+0x36a/0x1320
[  704.367294]  io_wq_submit_work+0x58/0x160
[  704.367295]  io_worker_handle_work+0x2a3/0x430
[  704.367296]  io_wqe_worker+0x2a0/0x350
[  704.367301]  kthread+0x136/0x180
[  704.367304]  ret_from_fork+0x22/0x30
---
 fs/io_uring.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/io_uring.c +1 -1
index 4e09af1d5d223..80a612d1e12bb 100644
@@ -645,12 +645,12 @@ struct io_kiocb {
 		 * restore the work, if needed.
 		 */
 		struct {
-			struct callback_head	task_work;
 			struct hlist_node	hash_node;
 			struct async_poll	*apoll;
 		};
 		struct io_wq_work	work;
 	};
+	struct callback_head	task_work;
 };
 
 #define IO_PLUG_THRESHOLD		2
Signature · 1 line
2.25.1
author maintainer
#9 lore raw

PATCH 8/60 fs/btrfs: Add cond_resched() for try_release_extent_mapping() stalls

In reply to Sasha Levin

From: "Paul E. McKenney" <[email protected]>

[ Upstream commit 9f47eb5461aaeb6cb8696f9d11503ae90e4d5cb0 ]

Very large I/Os can cause the following RCU CPU stall warning:

RIP: 0010:rb_prev+0x8/0x50
Code: 49 89 c0 49 89 d1 48 89 c2 48 89 f8 e9 e5 fd ff ff 4c 89 48 10 c3 4c =

89 06 c3 4c 89 40 10 c3 0f 1f 00 48 8b 0f 48 39 cf 74 38 <48> 8b 47 10 48 85 c0 74 22 48 8b 50 08 48 85 d2 74 0c 48 89 d0 48
RSP: 0018:ffffc9002212bab0 EFLAGS: 00000287 ORIG_RAX: ffffffffffffff13
RAX: ffff888821f93630 RBX: ffff888821f93630 RCX: ffff888821f937e0
RDX: 0000000000000000 RSI: 0000000000102000 RDI: ffff888821f93630
RBP: 0000000000103000 R08: 000000000006c000 R09: 0000000000000238
R10: 0000000000102fff R11: ffffc9002212bac8 R12: 0000000000000001
R13: ffffffffffffffff R14: 0000000000102000 R15: ffff888821f937e0

 __lookup_extent_mapping+0xa0/0x110
 try_release_extent_mapping+0xdc/0x220
 btrfs_releasepage+0x45/0x70
 shrink_page_list+0xa39/0xb30
 shrink_inactive_list+0x18f/0x3b0
 shrink_lruvec+0x38e/0x6b0
 shrink_node+0x14d/0x690
 do_try_to_free_pages+0xc6/0x3e0
 try_to_free_mem_cgroup_pages+0xe6/0x1e0
 reclaim_high.constprop.73+0x87/0xc0
 mem_cgroup_handle_over_high+0x66/0x150
 exit_to_usermode_loop+0x82/0xd0
 do_syscall_64+0xd4/0x100
 entry_SYSCALL_64_after_hwframe+0x44/0xa9

On a PREEMPT=n kernel, the try_release_extent_mapping() function's "while" loop might run for a very long time on a large I/O. This commit therefore adds a cond_resched() to this loop, providing RCU any needed quiescent states.

---
 fs/btrfs/extent_io.c | 2 ++
 1 file changed, 2 insertions(+)
fs/btrfs/extent_io.c +2 -0
index 79196eb1a1b36..9d6d646e1eb08 100644
@@ -4518,6 +4518,8 @@ int try_release_extent_mapping(struct page *page, gfp_t mask)
 
 			/* once for us */
 			free_extent_map(em);
+
+			cond_resched(); /* Allow large-extent preemption. */
 		}
 	}
 	return try_release_extent_state(tree, page, mask);
Signature · 1 line
2.25.1
author maintainer
#10 lore raw

PATCH 9/60 drm/amdgpu: avoid dereferencing a NULL pointer

In reply to Sasha Levin

From: Jack Xiao <[email protected]>

[ Upstream commit 55611b507fd6453d26030c0c0619fdf0c262766d ]

Check if irq_src is NULL to avoid dereferencing a NULL pointer, for MES ring is uneccessary to recieve an interrupt notification.

---
 drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c | 19 ++++++++++++-------
 1 file changed, 12 insertions(+), 7 deletions(-)
drivers/gpu/drm/amd/amdgpu/amdgpu_fence.c +12 -7
index 7531527067dfb..892c1e9a1eb04 100644
@@ -408,7 +408,9 @@ int amdgpu_fence_driver_start_ring(struct amdgpu_ring *ring,
 		ring->fence_drv.gpu_addr = adev->uvd.inst[ring->me].gpu_addr + index;
 	}
 	amdgpu_fence_write(ring, atomic_read(&ring->fence_drv.last_seq));
-	amdgpu_irq_get(adev, irq_src, irq_type);
+
+	if (irq_src)
+		amdgpu_irq_get(adev, irq_src, irq_type);
 
 	ring->fence_drv.irq_src = irq_src;
 	ring->fence_drv.irq_type = irq_type;
@@ -529,8 +531,9 @@ void amdgpu_fence_driver_fini(struct amdgpu_device *adev)
 			/* no need to trigger GPU reset as we are unloading */
 			amdgpu_fence_driver_force_completion(ring);
 		}
-		amdgpu_irq_put(adev, ring->fence_drv.irq_src,
-			       ring->fence_drv.irq_type);
+		if (ring->fence_drv.irq_src)
+			amdgpu_irq_put(adev, ring->fence_drv.irq_src,
+				       ring->fence_drv.irq_type);
 		drm_sched_fini(&ring->sched);
 		del_timer_sync(&ring->fence_drv.fallback_timer);
 		for (j = 0; j <= ring->fence_drv.num_fences_mask; ++j)
@@ -566,8 +569,9 @@ void amdgpu_fence_driver_suspend(struct amdgpu_device *adev)
 		}
 
 		/* disable the interrupt */
-		amdgpu_irq_put(adev, ring->fence_drv.irq_src,
-			       ring->fence_drv.irq_type);
+		if (ring->fence_drv.irq_src)
+			amdgpu_irq_put(adev, ring->fence_drv.irq_src,
+				       ring->fence_drv.irq_type);
 	}
 }
 
@@ -593,8 +597,9 @@ void amdgpu_fence_driver_resume(struct amdgpu_device *adev)
 			continue;
 
 		/* enable the interrupt */
-		amdgpu_irq_get(adev, ring->fence_drv.irq_src,
-			       ring->fence_drv.irq_type);
+		if (ring->fence_drv.irq_src)
+			amdgpu_irq_get(adev, ring->fence_drv.irq_src,
+				       ring->fence_drv.irq_type);
 	}
 }
 
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#11 lore raw

PATCH 10/60 drm/radeon: Fix reference count leaks caused by pm_runtime_get_sync

In reply to Sasha Levin

From: Aditya Pakki <[email protected]>

[ Upstream commit 9fb10671011143d15b6b40d6d5fa9c52c57e9d63 ]

On calling pm_runtime_get_sync() the reference count of the device is incremented. In case of failure, decrement the reference count before returning the error.

---
 drivers/gpu/drm/radeon/radeon_display.c | 4 +++-
 drivers/gpu/drm/radeon/radeon_drv.c     | 4 +++-
 drivers/gpu/drm/radeon/radeon_kms.c     | 4 +++-
 3 files changed, 9 insertions(+), 3 deletions(-)
drivers/gpu/drm/radeon/radeon_display.c +3 -1
index 35db79a168bf7..df1a7eb736517 100644
@@ -635,8 +635,10 @@ radeon_crtc_set_config(struct drm_mode_set *set,
 	dev = set->crtc->dev;
 
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0)
+	if (ret < 0) {
+		pm_runtime_put_autosuspend(dev->dev);
 		return ret;
+	}
 
 	ret = drm_crtc_helper_set_config(set, ctx);
 
drivers/gpu/drm/radeon/radeon_drv.c +3 -1
index 59f8186a24151..0cf0b00a0623e 100644
@@ -549,8 +549,10 @@ long radeon_drm_ioctl(struct file *filp,
 	long ret;
 	dev = file_priv->minor->dev;
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0)
+	if (ret < 0) {
+		pm_runtime_put_autosuspend(dev->dev);
 		return ret;
+	}
 
 	ret = drm_ioctl(filp, cmd, arg);
 	
drivers/gpu/drm/radeon/radeon_kms.c +3 -1
index 58176db85952c..779e4cd86245d 100644
@@ -638,8 +638,10 @@ int radeon_driver_open_kms(struct drm_device *dev, struct drm_file *file_priv)
 	file_priv->driver_priv = NULL;
 
 	r = pm_runtime_get_sync(dev->dev);
-	if (r < 0)
+	if (r < 0) {
+		pm_runtime_put_autosuspend(dev->dev);
 		return r;
+	}
 
 	/* new gpu have virtual address space support */
 	if (rdev->family >= CHIP_CAYMAN) {
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#12 lore raw

PATCH 11/60 crypto: aesni - Fix build with LLVM_IAS=1

In reply to Sasha Levin

From: Sedat Dilek <[email protected]>

[ Upstream commit 3347c8a079d67af21760a78cc5f2abbcf06d9571 ]

When building with LLVM_IAS=1 means using Clang's Integrated Assembly (IAS) from LLVM/Clang >= v10.0.1-rc1+ instead of GNU/as from GNU/binutils I see the following breakage in Debian/testing AMD64:

<instantiation>:15:74: error: too many positional arguments
PRECOMPUTE 8*3+8(%rsp), %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7,

                                                                         ^

arch/x86/crypto/aesni-intel_asm.S:1598:2: note: while in macro instantiation
GCM_INIT %r9, 8*3 +8(%rsp), 8*3 +16(%rsp), 8*3 +24(%rsp)
^
<instantiation>:47:2: error: unknown use of instruction mnemonic without a size suffix
GHASH_4_ENCRYPT_4_PARALLEL_dec %xmm9, %xmm10, %xmm11, %xmm12, %xmm13, %xmm14, %xmm0, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7, %xmm8, enc ^
arch/x86/crypto/aesni-intel_asm.S:1599:2: note: while in macro instantiation
GCM_ENC_DEC dec
^
<instantiation>:15:74: error: too many positional arguments
PRECOMPUTE 8*3+8(%rsp), %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7,

                                                                         ^

arch/x86/crypto/aesni-intel_asm.S:1686:2: note: while in macro instantiation
GCM_INIT %r9, 8*3 +8(%rsp), 8*3 +16(%rsp), 8*3 +24(%rsp)
^
<instantiation>:47:2: error: unknown use of instruction mnemonic without a size suffix
GHASH_4_ENCRYPT_4_PARALLEL_enc %xmm9, %xmm10, %xmm11, %xmm12, %xmm13, %xmm14, %xmm0, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7, %xmm8, enc ^
arch/x86/crypto/aesni-intel_asm.S:1687:2: note: while in macro instantiation
GCM_ENC_DEC enc

Craig Topper suggested me in ClangBuiltLinux issue #1050:

Quoted 7 lines

I think the "too many positional arguments" is because the parser isn't able to handle the trailing commas.

The "unknown use of instruction mnemonic" is because the macro was named GHASH_4_ENCRYPT_4_PARALLEL_DEC but its being instantiated with GHASH_4_ENCRYPT_4_PARALLEL_dec I guess gas ignores case on the macro instantiation, but llvm doesn't.

First, I removed the trailing comma in the PRECOMPUTE line.

Second, I substituted:
1. GHASH_4_ENCRYPT_4_PARALLEL_DEC -> GHASH_4_ENCRYPT_4_PARALLEL_dec
2. GHASH_4_ENCRYPT_4_PARALLEL_ENC -> GHASH_4_ENCRYPT_4_PARALLEL_enc

With these changes I was able to build with LLVM_IAS=1 and boot on bare metal.

I confirmed that this works with Linux-kernel v5.7.5 final.

NOTE: This patch is on top of Linux v5.7 final.

Thanks to Craig and especially Nick for double-checking and his comments.

---
 arch/x86/crypto/aesni-intel_asm.S | 6 +++---
 1 file changed, 3 insertions(+), 3 deletions(-)
arch/x86/crypto/aesni-intel_asm.S +3 -3
index cad6e1bfa7d5f..c216de2877426 100644
@@ -266,7 +266,7 @@ ALL_F:      .octa 0xffffffffffffffffffffffffffffffff
 	PSHUFB_XMM %xmm2, %xmm0
 	movdqu %xmm0, CurCount(%arg2) # ctx_data.current_counter = iv
 
-	PRECOMPUTE \SUBKEY, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7,
+	PRECOMPUTE \SUBKEY, %xmm1, %xmm2, %xmm3, %xmm4, %xmm5, %xmm6, %xmm7
 	movdqu HashKey(%arg2), %xmm13
 
 	CALC_AAD_HASH %xmm13, \AAD, \AADLEN, %xmm0, %xmm1, %xmm2, %xmm3, \
@@ -978,7 +978,7 @@ _initial_blocks_done\@:
 * arg1, %arg3, %arg4 are used as pointers only, not modified
 * %r11 is the data offset value
 */
-.macro GHASH_4_ENCRYPT_4_PARALLEL_ENC TMP1 TMP2 TMP3 TMP4 TMP5 \
+.macro GHASH_4_ENCRYPT_4_PARALLEL_enc TMP1 TMP2 TMP3 TMP4 TMP5 \
 TMP6 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 operation
 
 	movdqa	  \XMM1, \XMM5
@@ -1186,7 +1186,7 @@ aes_loop_par_enc_done\@:
 * arg1, %arg3, %arg4 are used as pointers only, not modified
 * %r11 is the data offset value
 */
-.macro GHASH_4_ENCRYPT_4_PARALLEL_DEC TMP1 TMP2 TMP3 TMP4 TMP5 \
+.macro GHASH_4_ENCRYPT_4_PARALLEL_dec TMP1 TMP2 TMP3 TMP4 TMP5 \
 TMP6 XMM0 XMM1 XMM2 XMM3 XMM4 XMM5 XMM6 XMM7 XMM8 operation
 
 	movdqa	  \XMM1, \XMM5
Signature · 1 line
2.25.1
author maintainer
#13 lore raw

PATCH 12/60 video: fbdev: savage: fix memory leak on error handling path in probe

In reply to Sasha Levin

From: Evgeny Novikov <[email protected]>

[ Upstream commit e8d35898a78e34fc854ed9680bc3f9caedab08cd ]

savagefb_probe() calls savage_init_fb_info() that can successfully allocate memory for info->pixmap.addr but then fail when fb_alloc_cmap() fails. savagefb_probe() goes to label failed_init and does not free allocated memory. It is not valid to go to label failed_mmio since savage_init_fb_info() can fail during memory allocation as well. So, the patch free allocated memory on the error handling path in savage_init_fb_info() itself.

Found by Linux Driver Verification project (linuxtesting.org).

---
 drivers/video/fbdev/savage/savagefb_driver.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/video/fbdev/savage/savagefb_driver.c +2 -0
index aab312a7d9da3..a542c33f20828 100644
@@ -2158,6 +2158,8 @@ static int savage_init_fb_info(struct fb_info *info, struct pci_dev *dev,
 			info->flags |= FBINFO_HWACCEL_COPYAREA |
 				       FBINFO_HWACCEL_FILLRECT |
 				       FBINFO_HWACCEL_IMAGEBLIT;
+		else
+			kfree(info->pixmap.addr);
 	}
 #endif
 	return err;
Signature · 1 line
2.25.1
author maintainer
#14 lore raw

PATCH 13/60 video: fbdev: neofb: fix memory leak in neo_scan_monitor()

In reply to Sasha Levin

From: Evgeny Novikov <[email protected]>

[ Upstream commit edcb3895a751c762a18d25c8d9846ce9759ed7e1 ]

neofb_probe() calls neo_scan_monitor() that can successfully allocate a memory for info->monspecs.modedb and proceed to case 0x03. There it does not free the memory and returns -1. neofb_probe() goes to label err_scan_monitor, thus, it does not free this memory through calling fb_destroy_modedb() as well. We can not go to label err_init_hw since neo_scan_monitor() can fail during memory allocation. So, the patch frees the memory directly for case 0x03.

Found by Linux Driver Verification project (linuxtesting.org).

---
 drivers/video/fbdev/neofb.c | 1 +
 1 file changed, 1 insertion(+)
drivers/video/fbdev/neofb.c +1 -0
index e6ea853c17238..5a363ce9b4cbe 100644
@@ -1820,6 +1820,7 @@ static int neo_scan_monitor(struct fb_info *info)
 #else
 		printk(KERN_ERR
 		       "neofb: Only 640x480, 800x600/480 and 1024x768 panels are currently supported\n");
+		kfree(info->monspecs.modedb);
 		return -1;
 #endif
 	default:
Signature · 1 line
2.25.1
author maintainer
#15 lore raw

PATCH 14/60 bus: ti-sysc: Add missing quirk flags for usb_host_hs

In reply to Sasha Levin

From: Tony Lindgren <[email protected]>

[ Upstream commit 4254632dba27271f6de66efd87e444ee405dee29 ]

Similar to what we have for the legacy platform data, we need to configure SWSUP_SIDLE and SWSUP_MSTANDBY quirks for usb_host_hs.

These are needed to drop the legacy platform data for usb_host_hs.

---
 drivers/bus/ti-sysc.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/bus/ti-sysc.c +4 -2
index 3b0417a014946..ae4cf4667633f 100644
@@ -1402,6 +1402,10 @@ static const struct sysc_revision_quirk sysc_revision_quirks[] = {
 		   SYSC_QUIRK_SWSUP_SIDLE | SYSC_QUIRK_SWSUP_MSTANDBY),
 	SYSC_QUIRK("tptc", 0, 0, -ENODEV, -ENODEV, 0x40007c00, 0xffffffff,
 		   SYSC_QUIRK_SWSUP_SIDLE | SYSC_QUIRK_SWSUP_MSTANDBY),
+	SYSC_QUIRK("usb_host_hs", 0, 0, 0x10, 0x14, 0x50700100, 0xffffffff,
+		   SYSC_QUIRK_SWSUP_SIDLE | SYSC_QUIRK_SWSUP_MSTANDBY),
+	SYSC_QUIRK("usb_host_hs", 0, 0, 0x10, -ENODEV, 0x50700101, 0xffffffff,
+		   SYSC_QUIRK_SWSUP_SIDLE | SYSC_QUIRK_SWSUP_MSTANDBY),
 	SYSC_QUIRK("usb_otg_hs", 0, 0x400, 0x404, 0x408, 0x00000050,
 		   0xffffffff, SYSC_QUIRK_SWSUP_SIDLE | SYSC_QUIRK_SWSUP_MSTANDBY),
 	SYSC_QUIRK("usb_otg_hs", 0, 0, 0x10, -ENODEV, 0x4ea2080d, 0xffffffff,
@@ -1473,8 +1477,6 @@ static const struct sysc_revision_quirk sysc_revision_quirks[] = {
 	SYSC_QUIRK("tpcc", 0, 0, -ENODEV, -ENODEV, 0x40014c00, 0xffffffff, 0),
 	SYSC_QUIRK("usbhstll", 0, 0, 0x10, 0x14, 0x00000004, 0xffffffff, 0),
 	SYSC_QUIRK("usbhstll", 0, 0, 0x10, 0x14, 0x00000008, 0xffffffff, 0),
-	SYSC_QUIRK("usb_host_hs", 0, 0, 0x10, 0x14, 0x50700100, 0xffffffff, 0),
-	SYSC_QUIRK("usb_host_hs", 0, 0, 0x10, -ENODEV, 0x50700101, 0xffffffff, 0),
 	SYSC_QUIRK("venc", 0x58003000, 0, -ENODEV, -ENODEV, 0x00000002, 0xffffffff, 0),
 	SYSC_QUIRK("vfpe", 0, 0, 0x104, -ENODEV, 0x4d001200, 0xffffffff, 0),
 #endif
Signature · 1 line
2.25.1
author maintainer
#16 lore raw

PATCH 15/60 md-cluster: fix wild pointer of unlock_all_bitmaps()

In reply to Sasha Levin

From: Zhao Heming <[email protected]>

[ Upstream commit 60f80d6f2d07a6d8aee485a1d1252327eeee0c81 ]

reproduction steps:

```
node1 # mdadm -C /dev/md0 -b clustered -e 1.2 -n 2 -l mirror /dev/sda
/dev/sdb
node2 # mdadm -A /dev/md0 /dev/sda /dev/sdb
node1 # mdadm -G /dev/md0 -b none
mdadm: failed to remove clustered bitmap.
node1 # mdadm -S --scan
^C  <==== mdadm hung & kernel crash
```

kernel stack:

```
[  335.230657] general protection fault: 0000 [#1] SMP NOPTI
[...]
[  335.230848] Call Trace:
[  335.230873]  ? unlock_all_bitmaps+0x5/0x70 [md_cluster]
[  335.230886]  unlock_all_bitmaps+0x3d/0x70 [md_cluster]
[  335.230899]  leave+0x10f/0x190 [md_cluster]
[  335.230932]  ? md_super_wait+0x93/0xa0 [md_mod]
[  335.230947]  ? leave+0x5/0x190 [md_cluster]
[  335.230973]  md_cluster_stop+0x1a/0x30 [md_mod]
[  335.230999]  md_bitmap_free+0x142/0x150 [md_mod]
[  335.231013]  ? _cond_resched+0x15/0x40
[  335.231025]  ? mutex_lock+0xe/0x30
[  335.231056]  __md_stop+0x1c/0xa0 [md_mod]
[  335.231083]  do_md_stop+0x160/0x580 [md_mod]
[  335.231119]  ? 0xffffffffc05fb078
[  335.231148]  md_ioctl+0xa04/0x1930 [md_mod]
[  335.231165]  ? filename_lookup+0xf2/0x190
[  335.231179]  blkdev_ioctl+0x93c/0xa10
[  335.231205]  ? _cond_resched+0x15/0x40
[  335.231214]  ? __check_object_size+0xd4/0x1a0
[  335.231224]  block_ioctl+0x39/0x40
[  335.231243]  do_vfs_ioctl+0xa0/0x680
[  335.231253]  ksys_ioctl+0x70/0x80
[  335.231261]  __x64_sys_ioctl+0x16/0x20
[  335.231271]  do_syscall_64+0x65/0x1f0
[  335.231278]  entry_SYSCALL_64_after_hwframe+0x44/0xa9
```
---
 drivers/md/md-cluster.c | 1 +
 1 file changed, 1 insertion(+)
drivers/md/md-cluster.c +1 -0
index 813a99ffa86f8..73fd50e779754 100644
@@ -1518,6 +1518,7 @@ static void unlock_all_bitmaps(struct mddev *mddev)
 			}
 		}
 		kfree(cinfo->other_bitmap_lockres);
+		cinfo->other_bitmap_lockres = NULL;
 	}
 }
 
Signature · 1 line
2.25.1
author maintainer
#17 lore raw

PATCH 16/60 drm/nouveau/kms/nv50-: Fix disabling dithering

In reply to Sasha Levin

From: Lyude Paul <[email protected]>

[ Upstream commit fb2420b701edbf96c2b6d557f0139902f455dc2b ]

While we expose the ability to turn off hardware dithering for nouveau, we actually make the mistake of turning it on anyway, due to dithering_depth containing a non-zero value if our dithering depth isn't also set to 6 bpc.

So, fix it by never enabling dithering when it's disabled.

---
 drivers/gpu/drm/nouveau/dispnv50/head.c | 24 +++++++++++++-----------
 1 file changed, 13 insertions(+), 11 deletions(-)
drivers/gpu/drm/nouveau/dispnv50/head.c +13 -11
index 8f6455697ba72..ed6819519f6d8 100644
@@ -84,18 +84,20 @@ nv50_head_atomic_check_dither(struct nv50_head_atom *armh,
 {
 	u32 mode = 0x00;
 
-	if (asyc->dither.mode == DITHERING_MODE_AUTO) {
-		if (asyh->base.depth > asyh->or.bpc * 3)
-			mode = DITHERING_MODE_DYNAMIC2X2;
-	} else {
-		mode = asyc->dither.mode;
-	}
+	if (asyc->dither.mode) {
+		if (asyc->dither.mode == DITHERING_MODE_AUTO) {
+			if (asyh->base.depth > asyh->or.bpc * 3)
+				mode = DITHERING_MODE_DYNAMIC2X2;
+		} else {
+			mode = asyc->dither.mode;
+		}
 
-	if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
-		if (asyh->or.bpc >= 8)
-			mode |= DITHERING_DEPTH_8BPC;
-	} else {
-		mode |= asyc->dither.depth;
+		if (asyc->dither.depth == DITHERING_DEPTH_AUTO) {
+			if (asyh->or.bpc >= 8)
+				mode |= DITHERING_DEPTH_8BPC;
+		} else {
+			mode |= asyc->dither.depth;
+		}
 	}
 
 	asyh->dither.enable = mode;
Signature · 1 line
2.25.1
author maintainer
#18 lore raw

PATCH 17/60 arm64: dts: hisilicon: hikey: fixes to comply with adi, adv7533 DT binding

In reply to Sasha Levin

From: Ricardo Cañuelo <[email protected]>

[ Upstream commit bbe28fc3cbabbef781bcdf847615d52ce2e26e42 ]

hi3660-hikey960.dts:
Define a 'ports' node for 'adv7533: adv7533@39' and the 'adi,dsi-lanes' property to make it compliant with the adi,adv7533 DT binding.

This fills the requirements to meet the binding requirements, remote endpoints are not defined.

hi6220-hikey.dts:
Change property name s/pd-gpio/pd-gpios, gpio properties should be plural. This is just a cosmetic change.

---
 arch/arm64/boot/dts/hisilicon/hi3660-hikey960.dts | 11 +++++++++++
 arch/arm64/boot/dts/hisilicon/hi6220-hikey.dts    |  2 +-
 2 files changed, 12 insertions(+), 1 deletion(-)
arch/arm64/boot/dts/hisilicon/hi3660-hikey960.dts +11 -0
index e035cf195b19a..8c4bfbaf3a806 100644
@@ -530,6 +530,17 @@ adv7533: adv7533@39 {
 		status = "ok";
 		compatible = "adi,adv7533";
 		reg = <0x39>;
+		adi,dsi-lanes = <4>;
+		ports {
+			#address-cells = <1>;
+			#size-cells = <0>;
+			port@0 {
+				reg = <0>;
+			};
+			port@1 {
+				reg = <1>;
+			};
+		};
 	};
 };
 
arch/arm64/boot/dts/hisilicon/hi6220-hikey.dts +1 -1
index c14205cd6bf5c..3e47150c05ec2 100644
@@ -516,7 +516,7 @@ adv7533: adv7533@39 {
 		reg = <0x39>;
 		interrupt-parent = <&gpio1>;
 		interrupts = <1 2>;
-		pd-gpio = <&gpio0 4 0>;
+		pd-gpios = <&gpio0 4 0>;
 		adi,dsi-lanes = <4>;
 		#sound-dai-cells = <0>;
 
Signature · 7 lines
2.25.1


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

PATCH 18/60 drm/etnaviv: fix ref count leak via pm_runtime_get_sync

In reply to Sasha Levin

From: Navid Emamdoost <[email protected]>

[ Upstream commit c5d5a32ead1e3a61a07a1e59eb52a53e4a6b2a7f ]

in etnaviv_gpu_submit, etnaviv_gpu_recover_hang, etnaviv_gpu_debugfs, and etnaviv_gpu_init the call to pm_runtime_get_sync increments the counter even in case of failure, leading to incorrect ref count. In case of failure, decrement the ref count before returning.

---
 drivers/gpu/drm/etnaviv/etnaviv_gpu.c | 14 ++++++++++----
 1 file changed, 10 insertions(+), 4 deletions(-)
drivers/gpu/drm/etnaviv/etnaviv_gpu.c +10 -4
index a31eeff2b297a..7c9f3f9ba1235 100644
@@ -722,7 +722,7 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu)
 	ret = pm_runtime_get_sync(gpu->dev);
 	if (ret < 0) {
 		dev_err(gpu->dev, "Failed to enable GPU power domain\n");
-		return ret;
+		goto pm_put;
 	}
 
 	etnaviv_hw_identify(gpu);
@@ -819,6 +819,7 @@ int etnaviv_gpu_init(struct etnaviv_gpu *gpu)
 
 fail:
 	pm_runtime_mark_last_busy(gpu->dev);
+pm_put:
 	pm_runtime_put_autosuspend(gpu->dev);
 
 	return ret;
@@ -859,7 +860,7 @@ int etnaviv_gpu_debugfs(struct etnaviv_gpu *gpu, struct seq_file *m)
 
 	ret = pm_runtime_get_sync(gpu->dev);
 	if (ret < 0)
-		return ret;
+		goto pm_put;
 
 	dma_lo = gpu_read(gpu, VIVS_FE_DMA_LOW);
 	dma_hi = gpu_read(gpu, VIVS_FE_DMA_HIGH);
@@ -1003,6 +1004,7 @@ int etnaviv_gpu_debugfs(struct etnaviv_gpu *gpu, struct seq_file *m)
 	ret = 0;
 
 	pm_runtime_mark_last_busy(gpu->dev);
+pm_put:
 	pm_runtime_put_autosuspend(gpu->dev);
 
 	return ret;
@@ -1016,7 +1018,7 @@ void etnaviv_gpu_recover_hang(struct etnaviv_gpu *gpu)
 	dev_err(gpu->dev, "recover hung GPU!\n");
 
 	if (pm_runtime_get_sync(gpu->dev) < 0)
-		return;
+		goto pm_put;
 
 	mutex_lock(&gpu->lock);
 
@@ -1035,6 +1037,7 @@ void etnaviv_gpu_recover_hang(struct etnaviv_gpu *gpu)
 
 	mutex_unlock(&gpu->lock);
 	pm_runtime_mark_last_busy(gpu->dev);
+pm_put:
 	pm_runtime_put_autosuspend(gpu->dev);
 }
 
@@ -1308,8 +1311,10 @@ struct dma_fence *etnaviv_gpu_submit(struct etnaviv_gem_submit *submit)
 
 	if (!submit->runtime_resumed) {
 		ret = pm_runtime_get_sync(gpu->dev);
-		if (ret < 0)
+		if (ret < 0) {
+			pm_runtime_put_noidle(gpu->dev);
 			return NULL;
+		}
 		submit->runtime_resumed = true;
 	}
 
@@ -1326,6 +1331,7 @@ struct dma_fence *etnaviv_gpu_submit(struct etnaviv_gem_submit *submit)
 	ret = event_alloc(gpu, nr_events, event);
 	if (ret) {
 		DRM_ERROR("no free events\n");
+		pm_runtime_put_noidle(gpu->dev);
 		return NULL;
 	}
 
Signature · 6 lines
2.25.1

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

PATCH 19/60 memory: samsung: exynos5422-dmc: Do not ignore return code of regmap_read()

In reply to Sasha Levin

From: Krzysztof Kozlowski <[email protected]>

[ Upstream commit c4f16e96d8fdd62ef12898fc0965c42093bed237 ]

Check for regmap_read() return code before using the read value in following write in exynos5_switch_timing_regs(). Pass reading error code to the callers.

This does not introduce proper error handling for such failed reads (and obviously regmap_write() error is still ignored) because the driver ignored this in all places. Therefor it only fixes reported issue while matching current driver coding style:

drivers/memory/samsung/exynos5422-dmc.c: In function 'exynos5_switch_timing_regs':
>> drivers/memory/samsung/exynos5422-dmc.c:216:6: warning: variable 'ret' set but not used [-Wunused-but-set-variable]

---
 drivers/memory/samsung/exynos5422-dmc.c | 12 +++++++++---
 1 file changed, 9 insertions(+), 3 deletions(-)
drivers/memory/samsung/exynos5422-dmc.c +9 -3
index 22a43d6628336..3460ba55fd596 100644
@@ -270,12 +270,14 @@ static int find_target_freq_idx(struct exynos5_dmc *dmc,
  * This function switches between these banks according to the
  * currently used clock source.
  */
-static void exynos5_switch_timing_regs(struct exynos5_dmc *dmc, bool set)
+static int exynos5_switch_timing_regs(struct exynos5_dmc *dmc, bool set)
 {
 	unsigned int reg;
 	int ret;
 
 	ret = regmap_read(dmc->clk_regmap, CDREX_LPDDR3PHY_CON3, &reg);
+	if (ret)
+		return ret;
 
 	if (set)
 		reg |= EXYNOS5_TIMING_SET_SWI;
@@ -283,6 +285,8 @@ static void exynos5_switch_timing_regs(struct exynos5_dmc *dmc, bool set)
 		reg &= ~EXYNOS5_TIMING_SET_SWI;
 
 	regmap_write(dmc->clk_regmap, CDREX_LPDDR3PHY_CON3, reg);
+
+	return 0;
 }
 
 /**
@@ -516,7 +520,7 @@ exynos5_dmc_switch_to_bypass_configuration(struct exynos5_dmc *dmc,
 	/*
 	 * Delays are long enough, so use them for the new coming clock.
 	 */
-	exynos5_switch_timing_regs(dmc, USE_MX_MSPLL_TIMINGS);
+	ret = exynos5_switch_timing_regs(dmc, USE_MX_MSPLL_TIMINGS);
 
 	return ret;
 }
@@ -577,7 +581,9 @@ exynos5_dmc_change_freq_and_volt(struct exynos5_dmc *dmc,
 
 	clk_set_rate(dmc->fout_bpll, target_rate);
 
-	exynos5_switch_timing_regs(dmc, USE_BPLL_TIMINGS);
+	ret = exynos5_switch_timing_regs(dmc, USE_BPLL_TIMINGS);
+	if (ret)
+		goto disable_clocks;
 
 	ret = clk_set_parent(dmc->mout_mclk_cdrex, dmc->mout_bpll);
 	if (ret)
Signature · 7 lines
2.25.1


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

PATCH 20/60 drm/nouveau: fix reference count leak in nouveau_debugfs_strap_peek

In reply to Sasha Levin

From: Aditya Pakki <[email protected]>

[ Upstream commit 8f29432417b11039ef960ab18987c7d61b2b5396 ]

nouveau_debugfs_strap_peek() calls pm_runtime_get_sync() that increments the reference count. In case of failure, decrement the ref count before returning the error.

---
 drivers/gpu/drm/nouveau/nouveau_debugfs.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/gpu/drm/nouveau/nouveau_debugfs.c +3 -1
index 15a3d40edf029..3e15a9d5e8faa 100644
@@ -54,8 +54,10 @@ nouveau_debugfs_strap_peek(struct seq_file *m, void *data)
 	int ret;
 
 	ret = pm_runtime_get_sync(drm->dev->dev);
-	if (ret < 0 && ret != -EACCES)
+	if (ret < 0 && ret != -EACCES) {
+		pm_runtime_put_autosuspend(drm->dev->dev);
 		return ret;
+	}
 
 	seq_printf(m, "0x%08x\n",
 		   nvif_rd32(&drm->client.device.object, 0x101000));
Signature · 1 line
2.25.1
author maintainer
#22 lore raw

PATCH 21/60 drm/nouveau: fix multiple instances of reference count leaks

In reply to Sasha Levin

From: Aditya Pakki <[email protected]>

[ Upstream commit 659fb5f154c3434c90a34586f3b7aa1c39cf6062 ]

On calling pm_runtime_get_sync() the reference count of the device is incremented. In case of failure, decrement the
ref count before returning the error.

---
 drivers/gpu/drm/nouveau/nouveau_drm.c | 8 ++++++--
 drivers/gpu/drm/nouveau/nouveau_gem.c | 4 +++-
 2 files changed, 9 insertions(+), 3 deletions(-)
drivers/gpu/drm/nouveau/nouveau_drm.c +6 -2
index ca4087f5a15b6..c484d21820c9b 100644
@@ -1051,8 +1051,10 @@ nouveau_drm_open(struct drm_device *dev, struct drm_file *fpriv)
 
 	/* need to bring up power immediately if opening device */
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0 && ret != -EACCES)
+	if (ret < 0 && ret != -EACCES) {
+		pm_runtime_put_autosuspend(dev->dev);
 		return ret;
+	}
 
 	get_task_comm(tmpname, current);
 	snprintf(name, sizeof(name), "%s[%d]", tmpname, pid_nr(fpriv->pid));
@@ -1134,8 +1136,10 @@ nouveau_drm_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
 	long ret;
 
 	ret = pm_runtime_get_sync(dev->dev);
-	if (ret < 0 && ret != -EACCES)
+	if (ret < 0 && ret != -EACCES) {
+		pm_runtime_put_autosuspend(dev->dev);
 		return ret;
+	}
 
 	switch (_IOC_NR(cmd) - DRM_COMMAND_BASE) {
 	case DRM_NOUVEAU_NVIF:
drivers/gpu/drm/nouveau/nouveau_gem.c +3 -1
index f5ece1f949734..f941ce8f81e3a 100644
@@ -45,8 +45,10 @@ nouveau_gem_object_del(struct drm_gem_object *gem)
 	int ret;
 
 	ret = pm_runtime_get_sync(dev);
-	if (WARN_ON(ret < 0 && ret != -EACCES))
+	if (WARN_ON(ret < 0 && ret != -EACCES)) {
+		pm_runtime_put_autosuspend(dev);
 		return;
+	}
 
 	if (gem->import_attach)
 		drm_prime_gem_destroy(gem, nvbo->bo.sg);
Signature · 1 line
2.25.1
author maintainer
#23 lore raw

PATCH 22/60 mmc: sdhci-cadence: do not use hardware tuning for SD mode

In reply to Sasha Levin

From: Masahiro Yamada <[email protected]>

[ Upstream commit adc40a5179df30421a5537bfeb4545100ab97d5e ]

As commit ef6b75671b5f ("mmc: sdhci-cadence: send tune request twice to work around errata") stated, this IP has an errata. This commit applies the second workaround for the SD mode.

Due to the errata, it is not possible to use the hardware tuning provided by SDHCI_HOST_CONTROL2.

Use the software-controlled tuning like the eMMC mode.

Set sdhci_host_ops::platform_execute_tuning instead of overriding mmc_host_ops::execute_tuning.

---
 drivers/mmc/host/sdhci-cadence.c | 123 ++++++++++++++++---------------
 1 file changed, 62 insertions(+), 61 deletions(-)
drivers/mmc/host/sdhci-cadence.c +62 -61
index 6da6d4fb5edd7..07a4cb989a68a 100644
@@ -194,57 +194,6 @@ static u32 sdhci_cdns_get_emmc_mode(struct sdhci_cdns_priv *priv)
 	return FIELD_GET(SDHCI_CDNS_HRS06_MODE, tmp);
 }
 
-static void sdhci_cdns_set_uhs_signaling(struct sdhci_host *host,
-					 unsigned int timing)
-{
-	struct sdhci_cdns_priv *priv = sdhci_cdns_priv(host);
-	u32 mode;
-
-	switch (timing) {
-	case MMC_TIMING_MMC_HS:
-		mode = SDHCI_CDNS_HRS06_MODE_MMC_SDR;
-		break;
-	case MMC_TIMING_MMC_DDR52:
-		mode = SDHCI_CDNS_HRS06_MODE_MMC_DDR;
-		break;
-	case MMC_TIMING_MMC_HS200:
-		mode = SDHCI_CDNS_HRS06_MODE_MMC_HS200;
-		break;
-	case MMC_TIMING_MMC_HS400:
-		if (priv->enhanced_strobe)
-			mode = SDHCI_CDNS_HRS06_MODE_MMC_HS400ES;
-		else
-			mode = SDHCI_CDNS_HRS06_MODE_MMC_HS400;
-		break;
-	default:
-		mode = SDHCI_CDNS_HRS06_MODE_SD;
-		break;
-	}
-
-	sdhci_cdns_set_emmc_mode(priv, mode);
-
-	/* For SD, fall back to the default handler */
-	if (mode == SDHCI_CDNS_HRS06_MODE_SD)
-		sdhci_set_uhs_signaling(host, timing);
-}
-
-static const struct sdhci_ops sdhci_cdns_ops = {
-	.set_clock = sdhci_set_clock,
-	.get_timeout_clock = sdhci_cdns_get_timeout_clock,
-	.set_bus_width = sdhci_set_bus_width,
-	.reset = sdhci_reset,
-	.set_uhs_signaling = sdhci_cdns_set_uhs_signaling,
-};
-
-static const struct sdhci_pltfm_data sdhci_cdns_uniphier_pltfm_data = {
-	.ops = &sdhci_cdns_ops,
-	.quirks2 = SDHCI_QUIRK2_PRESET_VALUE_BROKEN,
-};
-
-static const struct sdhci_pltfm_data sdhci_cdns_pltfm_data = {
-	.ops = &sdhci_cdns_ops,
-};
-
 static int sdhci_cdns_set_tune_val(struct sdhci_host *host, unsigned int val)
 {
 	struct sdhci_cdns_priv *priv = sdhci_cdns_priv(host);
@@ -278,23 +227,24 @@ static int sdhci_cdns_set_tune_val(struct sdhci_host *host, unsigned int val)
 	return 0;
 }
 
-static int sdhci_cdns_execute_tuning(struct mmc_host *mmc, u32 opcode)
+/*
+ * In SD mode, software must not use the hardware tuning and instead perform
+ * an almost identical procedure to eMMC.
+ */
+static int sdhci_cdns_execute_tuning(struct sdhci_host *host, u32 opcode)
 {
-	struct sdhci_host *host = mmc_priv(mmc);
 	int cur_streak = 0;
 	int max_streak = 0;
 	int end_of_streak = 0;
 	int i;
 
 	/*
-	 * This handler only implements the eMMC tuning that is specific to
-	 * this controller.  Fall back to the standard method for SD timing.
+	 * Do not execute tuning for UHS_SDR50 or UHS_DDR50.
+	 * The delay is set by probe, based on the DT properties.
 	 */
-	if (host->timing != MMC_TIMING_MMC_HS200)
-		return sdhci_execute_tuning(mmc, opcode);
-
-	if (WARN_ON(opcode != MMC_SEND_TUNING_BLOCK_HS200))
-		return -EINVAL;
+	if (host->timing != MMC_TIMING_MMC_HS200 &&
+	    host->timing != MMC_TIMING_UHS_SDR104)
+		return 0;
 
 	for (i = 0; i < SDHCI_CDNS_MAX_TUNING_LOOP; i++) {
 		if (sdhci_cdns_set_tune_val(host, i) ||
@@ -317,6 +267,58 @@ static int sdhci_cdns_execute_tuning(struct mmc_host *mmc, u32 opcode)
 	return sdhci_cdns_set_tune_val(host, end_of_streak - max_streak / 2);
 }
 
+static void sdhci_cdns_set_uhs_signaling(struct sdhci_host *host,
+					 unsigned int timing)
+{
+	struct sdhci_cdns_priv *priv = sdhci_cdns_priv(host);
+	u32 mode;
+
+	switch (timing) {
+	case MMC_TIMING_MMC_HS:
+		mode = SDHCI_CDNS_HRS06_MODE_MMC_SDR;
+		break;
+	case MMC_TIMING_MMC_DDR52:
+		mode = SDHCI_CDNS_HRS06_MODE_MMC_DDR;
+		break;
+	case MMC_TIMING_MMC_HS200:
+		mode = SDHCI_CDNS_HRS06_MODE_MMC_HS200;
+		break;
+	case MMC_TIMING_MMC_HS400:
+		if (priv->enhanced_strobe)
+			mode = SDHCI_CDNS_HRS06_MODE_MMC_HS400ES;
+		else
+			mode = SDHCI_CDNS_HRS06_MODE_MMC_HS400;
+		break;
+	default:
+		mode = SDHCI_CDNS_HRS06_MODE_SD;
+		break;
+	}
+
+	sdhci_cdns_set_emmc_mode(priv, mode);
+
+	/* For SD, fall back to the default handler */
+	if (mode == SDHCI_CDNS_HRS06_MODE_SD)
+		sdhci_set_uhs_signaling(host, timing);
+}
+
+static const struct sdhci_ops sdhci_cdns_ops = {
+	.set_clock = sdhci_set_clock,
+	.get_timeout_clock = sdhci_cdns_get_timeout_clock,
+	.set_bus_width = sdhci_set_bus_width,
+	.reset = sdhci_reset,
+	.platform_execute_tuning = sdhci_cdns_execute_tuning,
+	.set_uhs_signaling = sdhci_cdns_set_uhs_signaling,
+};
+
+static const struct sdhci_pltfm_data sdhci_cdns_uniphier_pltfm_data = {
+	.ops = &sdhci_cdns_ops,
+	.quirks2 = SDHCI_QUIRK2_PRESET_VALUE_BROKEN,
+};
+
+static const struct sdhci_pltfm_data sdhci_cdns_pltfm_data = {
+	.ops = &sdhci_cdns_ops,
+};
+
 static void sdhci_cdns_hs400_enhanced_strobe(struct mmc_host *mmc,
 					     struct mmc_ios *ios)
 {
@@ -377,7 +379,6 @@ static int sdhci_cdns_probe(struct platform_device *pdev)
 	priv->hrs_addr = host->ioaddr;
 	priv->enhanced_strobe = false;
 	host->ioaddr += SDHCI_CDNS_SRS_BASE;
-	host->mmc_host_ops.execute_tuning = sdhci_cdns_execute_tuning;
 	host->mmc_host_ops.hs400_enhanced_strobe =
 				sdhci_cdns_hs400_enhanced_strobe;
 	sdhci_enable_v4_mode(host);
Signature · 1 line
2.25.1
author maintainer
#24 lore raw

PATCH 23/60 btrfs: fix lockdep splat from btrfs_dump_space_info

In reply to Sasha Levin

From: Josef Bacik <[email protected]>

[ Upstream commit ab0db043c35da3477e57d4d516492b2d51a5ca0f ]

When running with -o enospc_debug you can get the following splat if one of the dump_space_info's trip

======================================================
WARNING: possible circular locking dependency detected
5.8.0-rc5+ #20 Tainted: G OE
------------------------------------------------------
dd/563090 is trying to acquire lock:
ffff9e7dbf4f1e18 (&ctl->tree_lock){+.+.}-{2:2}, at: btrfs_dump_free_space+0x2b/0xa0 [btrfs]

but task is already holding lock:
ffff9e7e2284d428 (&cache->lock){+.+.}-{2:2}, at: btrfs_dump_space_info+0xaa/0x120 [btrfs]

which lock already depends on the new lock.

the existing dependency chain (in reverse order) is:

-> #3 (&cache->lock){+.+.}-{2:2}:

	 _raw_spin_lock+0x25/0x30
	 btrfs_add_reserved_bytes+0x3c/0x3c0 [btrfs]
	 find_free_extent+0x7ef/0x13b0 [btrfs]
	 btrfs_reserve_extent+0x9b/0x180 [btrfs]
	 btrfs_alloc_tree_block+0xc1/0x340 [btrfs]
	 alloc_tree_block_no_bg_flush+0x4a/0x60 [btrfs]
	 __btrfs_cow_block+0x122/0x530 [btrfs]
	 btrfs_cow_block+0x106/0x210 [btrfs]
	 commit_cowonly_roots+0x55/0x300 [btrfs]
	 btrfs_commit_transaction+0x4ed/0xac0 [btrfs]
	 sync_filesystem+0x74/0x90
	 generic_shutdown_super+0x22/0x100
	 kill_anon_super+0x14/0x30
	 btrfs_kill_super+0x12/0x20 [btrfs]
	 deactivate_locked_super+0x36/0x70
	 cleanup_mnt+0x104/0x160
	 task_work_run+0x5f/0x90
	 __prepare_exit_to_usermode+0x1bd/0x1c0
	 do_syscall_64+0x5e/0xb0
	 entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> #2 (&space_info->lock){+.+.}-{2:2}:

	 _raw_spin_lock+0x25/0x30
	 btrfs_block_rsv_release+0x1a6/0x3f0 [btrfs]
	 btrfs_inode_rsv_release+0x4f/0x170 [btrfs]
	 btrfs_clear_delalloc_extent+0x155/0x480 [btrfs]
	 clear_state_bit+0x81/0x1a0 [btrfs]
	 __clear_extent_bit+0x25c/0x5d0 [btrfs]
	 clear_extent_bit+0x15/0x20 [btrfs]
	 btrfs_invalidatepage+0x2b7/0x3c0 [btrfs]
	 truncate_cleanup_page+0x47/0xe0
	 truncate_inode_pages_range+0x238/0x840
	 truncate_pagecache+0x44/0x60
	 btrfs_setattr+0x202/0x5e0 [btrfs]
	 notify_change+0x33b/0x490
	 do_truncate+0x76/0xd0
	 path_openat+0x687/0xa10
	 do_filp_open+0x91/0x100
	 do_sys_openat2+0x215/0x2d0
	 do_sys_open+0x44/0x80
	 do_syscall_64+0x52/0xb0
	 entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> #1 (&tree->lock#2){+.+.}-{2:2}:

	 _raw_spin_lock+0x25/0x30
	 find_first_extent_bit+0x32/0x150 [btrfs]
	 write_pinned_extent_entries.isra.0+0xc5/0x100 [btrfs]
	 __btrfs_write_out_cache+0x172/0x480 [btrfs]
	 btrfs_write_out_cache+0x7a/0xf0 [btrfs]
	 btrfs_write_dirty_block_groups+0x286/0x3b0 [btrfs]
	 commit_cowonly_roots+0x245/0x300 [btrfs]
	 btrfs_commit_transaction+0x4ed/0xac0 [btrfs]
	 close_ctree+0xf9/0x2f5 [btrfs]
	 generic_shutdown_super+0x6c/0x100
	 kill_anon_super+0x14/0x30
	 btrfs_kill_super+0x12/0x20 [btrfs]
	 deactivate_locked_super+0x36/0x70
	 cleanup_mnt+0x104/0x160
	 task_work_run+0x5f/0x90
	 __prepare_exit_to_usermode+0x1bd/0x1c0
	 do_syscall_64+0x5e/0xb0
	 entry_SYSCALL_64_after_hwframe+0x44/0xa9

-> #0 (&ctl->tree_lock){+.+.}-{2:2}:

	 __lock_acquire+0x1240/0x2460
	 lock_acquire+0xab/0x360
	 _raw_spin_lock+0x25/0x30
	 btrfs_dump_free_space+0x2b/0xa0 [btrfs]
	 btrfs_dump_space_info+0xf4/0x120 [btrfs]
	 btrfs_reserve_extent+0x176/0x180 [btrfs]
	 __btrfs_prealloc_file_range+0x145/0x550 [btrfs]
	 cache_save_setup+0x28d/0x3b0 [btrfs]
	 btrfs_start_dirty_block_groups+0x1fc/0x4f0 [btrfs]
	 btrfs_commit_transaction+0xcc/0xac0 [btrfs]
	 btrfs_alloc_data_chunk_ondemand+0x162/0x4c0 [btrfs]
	 btrfs_check_data_free_space+0x4c/0xa0 [btrfs]
	 btrfs_buffered_write.isra.0+0x19b/0x740 [btrfs]
	 btrfs_file_write_iter+0x3cf/0x610 [btrfs]
	 new_sync_write+0x11e/0x1b0
	 vfs_write+0x1c9/0x200
	 ksys_write+0x68/0xe0
	 do_syscall_64+0x52/0xb0
	 entry_SYSCALL_64_after_hwframe+0x44/0xa9

other info that might help us debug this:

  Chain exists of:
    &ctl->tree_lock --> &space_info->lock --> &cache->lock

   Possible unsafe locking scenario:
	 CPU0                    CPU1
	 ----                    ----
    lock(&cache->lock);
				 lock(&space_info->lock);
				 lock(&cache->lock);
    lock(&ctl->tree_lock);

*** DEADLOCK ***

6 locks held by dd/563090:
#0: ffff9e7e21d18448 (sb_writers#14){.+.+}-{0:0}, at: vfs_write+0x195/0x200
#1: ffff9e7dd0410ed8 (&sb->s_type->i_mutex_key#19){++++}-{3:3}, at: btrfs_file_write_iter+0x86/0x610 [btrfs] #2: ffff9e7e21d18638 (sb_internal#2){.+.+}-{0:0}, at: start_transaction+0x40b/0x5b0 [btrfs]
#3: ffff9e7e1f05d688 (&cur_trans->cache_write_mutex){+.+.}-{3:3}, at: btrfs_start_dirty_block_groups+0x158/0x4f0 [btrfs] #4: ffff9e7e2284ddb8 (&space_info->groups_sem){++++}-{3:3}, at: btrfs_dump_space_info+0x69/0x120 [btrfs] #5: ffff9e7e2284d428 (&cache->lock){+.+.}-{2:2}, at: btrfs_dump_space_info+0xaa/0x120 [btrfs]

  stack backtrace:
  CPU: 3 PID: 563090 Comm: dd Tainted: G           OE     5.8.0-rc5+ #20
  Hardware name: To Be Filled By O.E.M. To Be Filled By O.E.M./890FX Deluxe5, BIOS P1.40 05/03/2011
  Call Trace:
   dump_stack+0x96/0xd0
   check_noncircular+0x162/0x180
   __lock_acquire+0x1240/0x2460
   ? wake_up_klogd.part.0+0x30/0x40
   lock_acquire+0xab/0x360
   ? btrfs_dump_free_space+0x2b/0xa0 [btrfs]
   _raw_spin_lock+0x25/0x30
   ? btrfs_dump_free_space+0x2b/0xa0 [btrfs]
   btrfs_dump_free_space+0x2b/0xa0 [btrfs]
   btrfs_dump_space_info+0xf4/0x120 [btrfs]
   btrfs_reserve_extent+0x176/0x180 [btrfs]
   __btrfs_prealloc_file_range+0x145/0x550 [btrfs]
   ? btrfs_qgroup_reserve_data+0x1d/0x60 [btrfs]
   cache_save_setup+0x28d/0x3b0 [btrfs]
   btrfs_start_dirty_block_groups+0x1fc/0x4f0 [btrfs]
   btrfs_commit_transaction+0xcc/0xac0 [btrfs]
   ? start_transaction+0xe0/0x5b0 [btrfs]
   btrfs_alloc_data_chunk_ondemand+0x162/0x4c0 [btrfs]
   btrfs_check_data_free_space+0x4c/0xa0 [btrfs]
   btrfs_buffered_write.isra.0+0x19b/0x740 [btrfs]
   ? ktime_get_coarse_real_ts64+0xa8/0xd0
   ? trace_hardirqs_on+0x1c/0xe0
   btrfs_file_write_iter+0x3cf/0x610 [btrfs]
   new_sync_write+0x11e/0x1b0
   vfs_write+0x1c9/0x200
   ksys_write+0x68/0xe0
   do_syscall_64+0x52/0xb0
   entry_SYSCALL_64_after_hwframe+0x44/0xa9

This is because we're holding the block_group->lock while trying to dump the free space cache. However we don't need this lock, we just need it to read the values for the printk, so move the free space cache dumping outside of the block group lock.

---
 fs/btrfs/space-info.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/btrfs/space-info.c +1 -1
index 756950aba1a66..317d1d2160090 100644
@@ -468,8 +468,8 @@ void btrfs_dump_space_info(struct btrfs_fs_info *fs_info,
 			"block group %llu has %llu bytes, %llu used %llu pinned %llu reserved %s",
 			cache->start, cache->length, cache->used, cache->pinned,
 			cache->reserved, cache->ro ? "[readonly]" : "");
-		btrfs_dump_free_space(cache, bytes);
 		spin_unlock(&cache->lock);
+		btrfs_dump_free_space(cache, bytes);
 	}
 	if (++index < BTRFS_NR_RAID_TYPES)
 		goto again;
Signature · 1 line
2.25.1
author maintainer
#25 lore raw

PATCH 24/60 usb: mtu3: clear dual mode of u3port when disable device

In reply to Sasha Levin

From: Chunfeng Yun <[email protected]>

[ Upstream commit f1e51e99ed498d4aa9ae5df28e43d558ea627781 ]

If not clear u3port's dual mode when disable device, the IP will fail to enter sleep mode when suspend.

---
 drivers/usb/mtu3/mtu3_core.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/usb/mtu3/mtu3_core.c +5 -1
index 9dd02160cca97..e3780d4d65149 100644
@@ -131,8 +131,12 @@ static void mtu3_device_disable(struct mtu3 *mtu)
 	mtu3_setbits(ibase, SSUSB_U2_CTRL(0),
 		SSUSB_U2_PORT_DIS | SSUSB_U2_PORT_PDN);
 
-	if (mtu->ssusb->dr_mode == USB_DR_MODE_OTG)
+	if (mtu->ssusb->dr_mode == USB_DR_MODE_OTG) {
 		mtu3_clrbits(ibase, SSUSB_U2_CTRL(0), SSUSB_U2_PORT_OTG_SEL);
+		if (mtu->is_u3_ip)
+			mtu3_clrbits(ibase, SSUSB_U3_CTRL(0),
+				     SSUSB_U3_PORT_DUAL_MODE);
+	}
 
 	mtu3_setbits(ibase, U3D_SSUSB_IP_PW_CTRL2, SSUSB_IP_DEV_PDN);
 }
Signature · 7 lines
2.25.1


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

PATCH 25/60 drm: msm: a6xx: fix gpu failure after system resume

In reply to Sasha Levin

From: Akhil P Oommen <[email protected]>

[ Upstream commit 57c0bd517c06b088106b0236ed604056c8e06da5 ]

On targets where GMU is available, GMU takes over the ownership of GX GDSC during its initialization. So, move the refcount-get on GX PD before we initialize the GMU. This ensures that nobody can collapse the GX GDSC once GMU owns the GX GDSC. This patch fixes some GMU OOB errors seen during GPU wake up during a system resume.

---
 drivers/gpu/drm/msm/adreno/a6xx_gmu.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)
drivers/gpu/drm/msm/adreno/a6xx_gmu.c +10 -8
index 34607a98cc7c8..9a7a18951dc2b 100644
@@ -732,10 +732,19 @@ int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
 	/* Turn on the resources */
 	pm_runtime_get_sync(gmu->dev);
 
+	/*
+	 * "enable" the GX power domain which won't actually do anything but it
+	 * will make sure that the refcounting is correct in case we need to
+	 * bring down the GX after a GMU failure
+	 */
+	if (!IS_ERR_OR_NULL(gmu->gxpd))
+		pm_runtime_get_sync(gmu->gxpd);
+
 	/* Use a known rate to bring up the GMU */
 	clk_set_rate(gmu->core_clk, 200000000);
 	ret = clk_bulk_prepare_enable(gmu->nr_clocks, gmu->clocks);
 	if (ret) {
+		pm_runtime_put(gmu->gxpd);
 		pm_runtime_put(gmu->dev);
 		return ret;
 	}
@@ -771,19 +780,12 @@ int a6xx_gmu_resume(struct a6xx_gpu *a6xx_gpu)
 	/* Set the GPU to the current freq */
 	__a6xx_gmu_set_freq(gmu, gmu->current_perf_index);
 
-	/*
-	 * "enable" the GX power domain which won't actually do anything but it
-	 * will make sure that the refcounting is correct in case we need to
-	 * bring down the GX after a GMU failure
-	 */
-	if (!IS_ERR_OR_NULL(gmu->gxpd))
-		pm_runtime_get(gmu->gxpd);
-
 out:
 	/* On failure, shut down the GMU to leave it in a good state */
 	if (ret) {
 		disable_irq(gmu->gmu_irq);
 		a6xx_rpmh_stop(gmu);
+		pm_runtime_put(gmu->gxpd);
 		pm_runtime_put(gmu->dev);
 	}
 
Signature · 6 lines
2.25.1

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

PATCH 26/60 drm/msm: Fix a null pointer access in msm_gem_shrinker_count()

In reply to Sasha Levin

From: Akhil P Oommen <[email protected]>

[ Upstream commit 3cbdc8d8b7f39a7af3ea7b8dfa75caaebfda4e56 ]

Adding an msm_gem_object object to the inactive_list before completing its initialization is a bad idea because shrinker may pick it up from the inactive_list. Fix this by making sure that the initialization is complete before moving the msm_obj object to the inactive list.

This patch fixes the below error:

[10027.553044] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000068
[10027.573305] Mem abort info:
[10027.590160]   ESR = 0x96000006
[10027.597905]   EC = 0x25: DABT (current EL), IL = 32 bits
[10027.614430]   SET = 0, FnV = 0
[10027.624427]   EA = 0, S1PTW = 0
[10027.632722] Data abort info:
[10027.638039]   ISV = 0, ISS = 0x00000006
[10027.647459]   CM = 0, WnR = 0
[10027.654345] user pgtable: 4k pages, 39-bit VAs, pgdp=00000001e3a6a000
[10027.672681] [0000000000000068] pgd=0000000198c31003, pud=0000000198c31003, pmd=0000000000000000
[10027.693900] Internal error: Oops: 96000006 [#1] PREEMPT SMP
[10027.738261] CPU: 3 PID: 214 Comm: kswapd0 Tainted: G S                5.4.40 #1
[10027.745766] Hardware name: Qualcomm Technologies, Inc. SC7180 IDP (DT)
[10027.752472] pstate: 80c00009 (Nzcv daif +PAN +UAO)
[10027.757409] pc : mutex_is_locked+0x14/0x2c
[10027.761626] lr : msm_gem_shrinker_count+0x70/0xec
[10027.766454] sp : ffffffc011323ad0
[10027.769867] x29: ffffffc011323ad0 x28: ffffffe677e4b878
[10027.775324] x27: 0000000000000cc0 x26: 0000000000000000
[10027.780783] x25: ffffff817114a708 x24: 0000000000000008
[10027.786242] x23: ffffff8023ab7170 x22: 0000000000000001
[10027.791701] x21: ffffff817114a080 x20: 0000000000000119
[10027.797160] x19: 0000000000000068 x18: 00000000000003bc
[10027.802621] x17: 0000000004a34210 x16: 00000000000000c0
[10027.808083] x15: 0000000000000000 x14: 0000000000000000
[10027.813542] x13: ffffffe677e0a3c0 x12: 0000000000000000
[10027.819000] x11: 0000000000000000 x10: ffffff8174b94340
[10027.824461] x9 : 0000000000000000 x8 : 0000000000000000
[10027.829919] x7 : 00000000000001fc x6 : ffffffc011323c88
[10027.835373] x5 : 0000000000000001 x4 : ffffffc011323d80
[10027.840832] x3 : ffffffff0477b348 x2 : 0000000000000000
[10027.846290] x1 : ffffffc011323b68 x0 : 0000000000000068
[10027.851748] Call trace:
[10027.854264]  mutex_is_locked+0x14/0x2c
[10027.858121]  msm_gem_shrinker_count+0x70/0xec
[10027.862603]  shrink_slab+0xc0/0x4b4
[10027.866187]  shrink_node+0x4a8/0x818
[10027.869860]  kswapd+0x624/0x890
[10027.873097]  kthread+0x11c/0x12c
[10027.876424]  ret_from_fork+0x10/0x18
[10027.880102] Code: f9000bf3 910003fd aa0003f3 d503201f (f9400268)
[10027.886362] ---[ end trace df5849a1a3543251 ]---
[10027.891518] Kernel panic - not syncing: Fatal exception
---
 drivers/gpu/drm/msm/msm_gem.c | 36 ++++++++++++++++++++---------------
 1 file changed, 21 insertions(+), 15 deletions(-)
drivers/gpu/drm/msm/msm_gem.c +21 -15
index 5a6a79fbc9d6e..d92a0ffe2a767 100644
@@ -977,10 +977,8 @@ int msm_gem_new_handle(struct drm_device *dev, struct drm_file *file,
 
 static int msm_gem_new_impl(struct drm_device *dev,
 		uint32_t size, uint32_t flags,
-		struct drm_gem_object **obj,
-		bool struct_mutex_locked)
+		struct drm_gem_object **obj)
 {
-	struct msm_drm_private *priv = dev->dev_private;
 	struct msm_gem_object *msm_obj;
 
 	switch (flags & MSM_BO_CACHE_MASK) {
@@ -1006,15 +1004,6 @@ static int msm_gem_new_impl(struct drm_device *dev,
 	INIT_LIST_HEAD(&msm_obj->submit_entry);
 	INIT_LIST_HEAD(&msm_obj->vmas);
 
-	if (struct_mutex_locked) {
-		WARN_ON(!mutex_is_locked(&dev->struct_mutex));
-		list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
-	} else {
-		mutex_lock(&dev->struct_mutex);
-		list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
-		mutex_unlock(&dev->struct_mutex);
-	}
-
 	*obj = &msm_obj->base;
 
 	return 0;
@@ -1024,6 +1013,7 @@ static struct drm_gem_object *_msm_gem_new(struct drm_device *dev,
 		uint32_t size, uint32_t flags, bool struct_mutex_locked)
 {
 	struct msm_drm_private *priv = dev->dev_private;
+	struct msm_gem_object *msm_obj;
 	struct drm_gem_object *obj = NULL;
 	bool use_vram = false;
 	int ret;
@@ -1044,14 +1034,15 @@ static struct drm_gem_object *_msm_gem_new(struct drm_device *dev,
 	if (size == 0)
 		return ERR_PTR(-EINVAL);
 
-	ret = msm_gem_new_impl(dev, size, flags, &obj, struct_mutex_locked);
+	ret = msm_gem_new_impl(dev, size, flags, &obj);
 	if (ret)
 		goto fail;
 
+	msm_obj = to_msm_bo(obj);
+
 	if (use_vram) {
 		struct msm_gem_vma *vma;
 		struct page **pages;
-		struct msm_gem_object *msm_obj = to_msm_bo(obj);
 
 		mutex_lock(&msm_obj->lock);
 
@@ -1086,6 +1077,15 @@ static struct drm_gem_object *_msm_gem_new(struct drm_device *dev,
 		mapping_set_gfp_mask(obj->filp->f_mapping, GFP_HIGHUSER);
 	}
 
+	if (struct_mutex_locked) {
+		WARN_ON(!mutex_is_locked(&dev->struct_mutex));
+		list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
+	} else {
+		mutex_lock(&dev->struct_mutex);
+		list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
+		mutex_unlock(&dev->struct_mutex);
+	}
+
 	return obj;
 
 fail:
@@ -1108,6 +1108,7 @@ struct drm_gem_object *msm_gem_new(struct drm_device *dev,
 struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 		struct dma_buf *dmabuf, struct sg_table *sgt)
 {
+	struct msm_drm_private *priv = dev->dev_private;
 	struct msm_gem_object *msm_obj;
 	struct drm_gem_object *obj;
 	uint32_t size;
@@ -1121,7 +1122,7 @@ struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 
 	size = PAGE_ALIGN(dmabuf->size);
 
-	ret = msm_gem_new_impl(dev, size, MSM_BO_WC, &obj, false);
+	ret = msm_gem_new_impl(dev, size, MSM_BO_WC, &obj);
 	if (ret)
 		goto fail;
 
@@ -1146,6 +1147,11 @@ struct drm_gem_object *msm_gem_import(struct drm_device *dev,
 	}
 
 	mutex_unlock(&msm_obj->lock);
+
+	mutex_lock(&dev->struct_mutex);
+	list_add_tail(&msm_obj->mm_list, &priv->inactive_list);
+	mutex_unlock(&dev->struct_mutex);
+
 	return obj;
 
 fail:
Signature · 6 lines
2.25.1

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

PATCH 27/60 drm/debugfs: fix plain echo to connector "force" attribute

In reply to Sasha Levin

From: Michael Tretter <[email protected]>

[ Upstream commit c704b17071c4dc571dca3af4e4151dac51de081a ]

Using plain echo to set the "force" connector attribute fails with -EINVAL, because echo appends a newline to the output.

Replace strcmp with sysfs_streq to also accept strings that end with a newline.

v2: use sysfs_streq instead of stripping trailing whitespace

---
 drivers/gpu/drm/drm_debugfs.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/gpu/drm/drm_debugfs.c +4 -4
index 4e673d318503c..fb251c00fdd3e 100644
@@ -336,13 +336,13 @@ static ssize_t connector_write(struct file *file, const char __user *ubuf,
 
 	buf[len] = '\0';
 
-	if (!strcmp(buf, "on"))
+	if (sysfs_streq(buf, "on"))
 		connector->force = DRM_FORCE_ON;
-	else if (!strcmp(buf, "digital"))
+	else if (sysfs_streq(buf, "digital"))
 		connector->force = DRM_FORCE_ON_DIGITAL;
-	else if (!strcmp(buf, "off"))
+	else if (sysfs_streq(buf, "off"))
 		connector->force = DRM_FORCE_OFF;
-	else if (!strcmp(buf, "unspecified"))
+	else if (sysfs_streq(buf, "unspecified"))
 		connector->force = DRM_FORCE_UNSPECIFIED;
 	else
 		return -EINVAL;
Signature · 6 lines
2.25.1

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

PATCH 28/60 drm/radeon: disable AGP by default

In reply to Sasha Levin

From: Christian König <[email protected]>

[ Upstream commit ba806f98f868ce107aa9c453fef751de9980e4af ]

Always use the PCI GART instead. We just have to many cases where AGP still causes problems. This means a performance regression for some GPUs, but also a bug fix for some others.

---
 drivers/gpu/drm/radeon/radeon_drv.c | 5 -----
 1 file changed, 5 deletions(-)
drivers/gpu/drm/radeon/radeon_drv.c +0 -5
index 0cf0b00a0623e..6f0d1971099b7 100644
@@ -171,12 +171,7 @@ int radeon_no_wb;
 int radeon_modeset = -1;
 int radeon_dynclks = -1;
 int radeon_r4xx_atom = 0;
-#ifdef __powerpc__
-/* Default to PCI on PowerPC (fdo #95017) */
 int radeon_agpmode = -1;
-#else
-int radeon_agpmode = 0;
-#endif
 int radeon_vram_limit = 0;
 int radeon_gart_size = -1; /* auto */
 int radeon_benchmarking = 0;
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#30 lore raw

PATCH 29/60 net: phy: mscc: restore the base page in vsc8514/8584_config_init

In reply to Sasha Levin

From: Antoine Tenart <[email protected]>

[ Upstream commit 6119dda34e5d0821959e37641b287576826b6378 ]

In the vsc8584_config_init and vsc8514_config_init, the base page is set to 'GPIO', configuration is done, and the page is never explicitly restored to the standard page. No bug was triggered as it turns out helpers called in those config_init functions do modify the base page, and set it back to standard. But that is dangerous and any modification to those functions would introduce bugs. This patch fixes this, to improve maintenance, by restoring the base page to 'standard' once 'GPIO' accesses are completed.

---
 drivers/net/phy/mscc/mscc_main.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/net/phy/mscc/mscc_main.c +9 -0
index 485a4f8a6a9a6..95bd2d277ba42 100644
@@ -1413,6 +1413,11 @@ static int vsc8584_config_init(struct phy_device *phydev)
 	if (ret)
 		goto err;
 
+	ret = phy_base_write(phydev, MSCC_EXT_PAGE_ACCESS,
+			     MSCC_PHY_PAGE_STANDARD);
+	if (ret)
+		goto err;
+
 	if (!phy_interface_is_rgmii(phydev)) {
 		val = PROC_CMD_MCB_ACCESS_MAC_CONF | PROC_CMD_RST_CONF_PORT |
 			PROC_CMD_READ_MOD_WRITE_PORT;
@@ -1799,7 +1804,11 @@ static int vsc8514_config_init(struct phy_device *phydev)
 	val &= ~MAC_CFG_MASK;
 	val |= MAC_CFG_QSGMII;
 	ret = phy_base_write(phydev, MSCC_PHY_MAC_CFG_FASTLINK, val);
+	if (ret)
+		goto err;
 
+	ret = phy_base_write(phydev, MSCC_EXT_PAGE_ACCESS,
+			     MSCC_PHY_PAGE_STANDARD);
 	if (ret)
 		goto err;
 
Signature · 1 line
2.25.1
author maintainer
#31 lore raw

PATCH 30/60 irqchip/irq-mtk-sysirq: Replace spinlock with raw_spinlock

In reply to Sasha Levin

From: Bartosz Golaszewski <[email protected]>

[ Upstream commit 6eeb997ab5075e770a002c51351fa4ec2c6b5c39 ]

This driver may take a regular spinlock when a raw spinlock (irq_desc->lock) is already taken which results in the following lockdep splat:

=============================
[ BUG: Invalid wait context ]
5.7.0-rc7 #1 Not tainted
-----------------------------
swapper/0/0 is trying to lock:
ffffff800303b798 (&chip_data->lock){....}-{3:3}, at: mtk_sysirq_set_type+0x48/0xc0 other info that might help us debug this:
context-{5:5}
2 locks held by swapper/0/0:
#0: ffffff800302ee68 (&desc->request_mutex){....}-{4:4}, at: __setup_irq+0xc4/0x8a0 #1: ffffff800302ecf0 (&irq_desc_lock_class){....}-{2:2}, at: __setup_irq+0xe4/0x8a0

stack backtrace:
CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.7.0-rc7 #1
Hardware name: Pumpkin MT8516 (DT)
Call trace:
 dump_backtrace+0x0/0x180
 show_stack+0x14/0x20
 dump_stack+0xd0/0x118
 __lock_acquire+0x8c8/0x2270
 lock_acquire+0xf8/0x470
 _raw_spin_lock_irqsave+0x50/0x78
 mtk_sysirq_set_type+0x48/0xc0
 __irq_set_trigger+0x58/0x170
 __setup_irq+0x420/0x8a0
 request_threaded_irq+0xd8/0x190
 timer_of_init+0x1e8/0x2c4
 mtk_gpt_init+0x5c/0x1dc
 timer_probe+0x74/0xf4
 time_init+0x14/0x44
 start_kernel+0x394/0x4f0

Replace the spinlock_t with raw_spinlock_t to avoid this warning.

---
 drivers/irqchip/irq-mtk-sysirq.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)
drivers/irqchip/irq-mtk-sysirq.c +4 -4
index 73eae5966a403..6ff98b87e5c04 100644
@@ -15,7 +15,7 @@
 #include <linux/spinlock.h>
 
 struct mtk_sysirq_chip_data {
-	spinlock_t lock;
+	raw_spinlock_t lock;
 	u32 nr_intpol_bases;
 	void __iomem **intpol_bases;
 	u32 *intpol_words;
@@ -37,7 +37,7 @@ static int mtk_sysirq_set_type(struct irq_data *data, unsigned int type)
 	reg_index = chip_data->which_word[hwirq];
 	offset = hwirq & 0x1f;
 
-	spin_lock_irqsave(&chip_data->lock, flags);
+	raw_spin_lock_irqsave(&chip_data->lock, flags);
 	value = readl_relaxed(base + reg_index * 4);
 	if (type == IRQ_TYPE_LEVEL_LOW || type == IRQ_TYPE_EDGE_FALLING) {
 		if (type == IRQ_TYPE_LEVEL_LOW)
@@ -53,7 +53,7 @@ static int mtk_sysirq_set_type(struct irq_data *data, unsigned int type)
 
 	data = data->parent_data;
 	ret = data->chip->irq_set_type(data, type);
-	spin_unlock_irqrestore(&chip_data->lock, flags);
+	raw_spin_unlock_irqrestore(&chip_data->lock, flags);
 	return ret;
 }
 
@@ -212,7 +212,7 @@ static int __init mtk_sysirq_of_init(struct device_node *node,
 		ret = -ENOMEM;
 		goto out_free_which_word;
 	}
-	spin_lock_init(&chip_data->lock);
+	raw_spin_lock_init(&chip_data->lock);
 
 	return 0;
 
Signature · 1 line
2.25.1
author maintainer
#32 lore raw

PATCH 31/60 mm/mmap.c: Add cond_resched() for exit_mmap() CPU stalls

In reply to Sasha Levin

From: "Paul E. McKenney" <[email protected]>

[ Upstream commit 0a3b3c253a1eb2c7fe7f34086d46660c909abeb3 ]

A large process running on a heavily loaded system can encounter the following RCU CPU stall warning:

  rcu: INFO: rcu_sched self-detected stall on CPU
  rcu: 	3-....: (20998 ticks this GP) idle=4ea/1/0x4000000000000002 softirq=556558/556558 fqs=5190
  	(t=21013 jiffies g=1005461 q=132576)
  NMI backtrace for cpu 3
  CPU: 3 PID: 501900 Comm: aio-free-ring-w Kdump: loaded Not tainted 5.2.9-108_fbk12_rc3_3858_gb83b75af7909 #1
  Hardware name: Wiwynn   HoneyBadger/PantherPlus, BIOS HBM6.71 02/03/2016
  Call Trace:
   <IRQ>
   dump_stack+0x46/0x60
   nmi_cpu_backtrace.cold.3+0x13/0x50
   ? lapic_can_unplug_cpu.cold.27+0x34/0x34
   nmi_trigger_cpumask_backtrace+0xba/0xca
   rcu_dump_cpu_stacks+0x99/0xc7
   rcu_sched_clock_irq.cold.87+0x1aa/0x397
   ? tick_sched_do_timer+0x60/0x60
   update_process_times+0x28/0x60
   tick_sched_timer+0x37/0x70
   __hrtimer_run_queues+0xfe/0x270
   hrtimer_interrupt+0xf4/0x210
   smp_apic_timer_interrupt+0x5e/0x120
   apic_timer_interrupt+0xf/0x20
   </IRQ>
  RIP: 0010:kmem_cache_free+0x223/0x300
  Code: 88 00 00 00 0f 85 ca 00 00 00 41 8b 55 18 31 f6 f7 da 41 f6 45 0a 02 40 0f 94 c6 83 c6 05 9c 41 5e fa e8 a0 a7 01 00 41 56 9d <49> 8b 47 08 a8 03 0f 85 87 00 00 00 65 48 ff 08 e9 3d fe ff ff 65
  RSP: 0018:ffffc9000e8e3da8 EFLAGS: 00000206 ORIG_RAX: ffffffffffffff13
  RAX: 0000000000020000 RBX: ffff88861b9de960 RCX: 0000000000000030
  RDX: fffffffffffe41e8 RSI: 000060777fe3a100 RDI: 000000000001be18
  RBP: ffffea00186e7780 R08: ffffffffffffffff R09: ffffffffffffffff
  R10: ffff88861b9dea28 R11: ffff88887ffde000 R12: ffffffff81230a1f
  R13: ffff888854684dc0 R14: 0000000000000206 R15: ffff8888547dbc00
   ? remove_vma+0x4f/0x60
   remove_vma+0x4f/0x60
   exit_mmap+0xd6/0x160
   mmput+0x4a/0x110
   do_exit+0x278/0xae0
   ? syscall_trace_enter+0x1d3/0x2b0
   ? handle_mm_fault+0xaa/0x1c0
   do_group_exit+0x3a/0xa0
   __x64_sys_exit_group+0x14/0x20
   do_syscall_64+0x42/0x100
   entry_SYSCALL_64_after_hwframe+0x44/0xa9

And on a PREEMPT=n kernel, the "while (vma)" loop in exit_mmap() can run for a very long time given a large process. This commit therefore adds a cond_resched() to this loop, providing RCU any needed quiescent states.

---
 mm/mmap.c | 1 +
 1 file changed, 1 insertion(+)
mm/mmap.c +1 -0
index bb1822ac99090..55bb456fd0d0f 100644
@@ -3171,6 +3171,7 @@ void exit_mmap(struct mm_struct *mm)
 		if (vma->vm_flags & VM_ACCOUNT)
 			nr_accounted += vma_pages(vma);
 		vma = remove_vma(vma);
+		cond_resched();
 	}
 	vm_unacct_memory(nr_accounted);
 }
Signature · 1 line
2.25.1
author maintainer
#33 lore raw

PATCH 32/60 drm/amd/display: Improve DisplayPort monitor interop

In reply to Sasha Levin

From: Aric Cyr <[email protected]>

[ Upstream commit eec3303de3378cdfaa0bb86f43546dbbd88f94e2 ]

[Why]
DC is very fast at link training and stream enablement which causes issues such as blackscreens for non-compliant monitors.

[How]
After debugging with scaler vendors we implement the minimum delays at the necessary locations to ensure the monitor does not hang. Delays are generic due to lack of IEEE OUI information on the failing displays.

---
 drivers/gpu/drm/amd/display/dc/core/dc_link.c    |  4 +++-
 drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c | 16 ++++++++++------
 .../amd/display/dc/dce110/dce110_hw_sequencer.c  | 11 ++++++++++-
 3 files changed, 23 insertions(+), 8 deletions(-)
drivers/gpu/drm/amd/display/dc/core/dc_link.c +3 -1
index 67cfff1586e9f..3f157bcc174b9 100644
@@ -3146,9 +3146,11 @@ void core_link_disable_stream(struct pipe_ctx *pipe_ctx)
 			write_i2c_redriver_setting(pipe_ctx, false);
 		}
 	}
-	dc->hwss.disable_stream(pipe_ctx);
 
 	disable_link(pipe_ctx->stream->link, pipe_ctx->stream->signal);
+
+	dc->hwss.disable_stream(pipe_ctx);
+
 	if (pipe_ctx->stream->timing.flags.DSC) {
 		if (dc_is_dp_signal(pipe_ctx->stream->signal))
 			dp_set_dsc_enable(pipe_ctx, false);
drivers/gpu/drm/amd/display/dc/core/dc_link_dp.c +10 -6
index caa090d0b6acc..1ada01322cd2c 100644
@@ -1103,6 +1103,10 @@ static inline enum link_training_result perform_link_training_int(
 	dpcd_pattern.v1_4.TRAINING_PATTERN_SET = DPCD_TRAINING_PATTERN_VIDEOIDLE;
 	dpcd_set_training_pattern(link, dpcd_pattern);
 
+	/* delay 5ms after notifying sink of idle pattern before switching output */
+	if (link->connector_signal != SIGNAL_TYPE_EDP)
+		msleep(5);
+
 	/* 4. mainlink output idle pattern*/
 	dp_set_hw_test_pattern(link, DP_TEST_PATTERN_VIDEO_MODE, NULL, 0);
 
@@ -1552,6 +1556,12 @@ bool perform_link_training_with_retries(
 	struct dc_link *link = stream->link;
 	enum dp_panel_mode panel_mode = dp_get_panel_mode(link);
 
+	/* We need to do this before the link training to ensure the idle pattern in SST
+	 * mode will be sent right after the link training
+	 */
+	link->link_enc->funcs->connect_dig_be_to_fe(link->link_enc,
+							pipe_ctx->stream_res.stream_enc->id, true);
+
 	for (j = 0; j < attempts; ++j) {
 
 		dp_enable_link_phy(
@@ -1568,12 +1578,6 @@ bool perform_link_training_with_retries(
 
 		dp_set_panel_mode(link, panel_mode);
 
-		/* We need to do this before the link training to ensure the idle pattern in SST
-		 * mode will be sent right after the link training
-		 */
-		link->link_enc->funcs->connect_dig_be_to_fe(link->link_enc,
-								pipe_ctx->stream_res.stream_enc->id, true);
-
 		if (link->aux_access_disabled) {
 			dc_link_dp_perform_link_training_skip_aux(link, link_setting);
 			return true;
drivers/gpu/drm/amd/display/dc/dce110/dce110_hw_sequencer.c +10 -1
index 10527593868cc..24ca592c90df5 100644
@@ -1090,8 +1090,17 @@ void dce110_blank_stream(struct pipe_ctx *pipe_ctx)
 		dc_link_set_abm_disable(link);
 	}
 
-	if (dc_is_dp_signal(pipe_ctx->stream->signal))
+	if (dc_is_dp_signal(pipe_ctx->stream->signal)) {
 		pipe_ctx->stream_res.stream_enc->funcs->dp_blank(pipe_ctx->stream_res.stream_enc);
+
+		/*
+		 * After output is idle pattern some sinks need time to recognize the stream
+		 * has changed or they enter protection state and hang.
+		 */
+		if (!dc_is_embedded_signal(pipe_ctx->stream->signal))
+			msleep(60);
+	}
+
 }
 
 
Signature · 6 lines
2.25.1

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

PATCH 33/60 drm/amdgpu/display bail early in dm_pp_get_static_clocks

In reply to Sasha Levin

From: Alex Deucher <[email protected]>

[ Upstream commit 376814f5fcf1aadda501d1413d56e8af85d19a97 ]

If there are no supported callbacks. We'll fall back to the nominal clocks.

---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c +2 -0
index a2e1a73f66b81..7cee8070cb113 100644
@@ -530,6 +530,8 @@ bool dm_pp_get_static_clocks(
 			&pp_clk_info);
 	else if (adev->smu.ppt_funcs)
 		ret = smu_get_current_clocks(&adev->smu, &pp_clk_info);
+	else
+		return false;
 	if (ret)
 		return false;
 
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#35 lore raw

PATCH 34/60 drm/amdgpu/display: properly guard the calls to swSMU functions

In reply to Sasha Levin

From: Alex Deucher <[email protected]>

[ Upstream commit 4072327a2622af8688b88f5cd0a472136d3bf33d ]

It's only applicable on newer asics. We could end up here when using DC on older asics like SI or KV.

---
 drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/display/amdgpu_dm/amdgpu_dm_pp_smu.c +2 -2
index 7cee8070cb113..5c6a6ae48d396 100644
@@ -106,7 +106,7 @@ bool dm_pp_apply_display_requirements(
 			adev->powerplay.pp_funcs->display_configuration_change(
 				adev->powerplay.pp_handle,
 				&adev->pm.pm_display_cfg);
-		else
+		else if (adev->smu.ppt_funcs)
 			smu_display_configuration_change(smu,
 							 &adev->pm.pm_display_cfg);
 
@@ -592,7 +592,7 @@ void pp_rv_set_wm_ranges(struct pp_smu *pp,
 	if (pp_funcs && pp_funcs->set_watermarks_for_clocks_ranges)
 		pp_funcs->set_watermarks_for_clocks_ranges(pp_handle,
 							   &wm_with_clock_ranges);
-	else
+	else if (adev->smu.ppt_funcs)
 		smu_set_watermarks_for_clock_ranges(&adev->smu,
 				&wm_with_clock_ranges);
 }
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#36 lore raw

PATCH 35/60 drm/amd/powerplay: fix compile error with ARCH=arc

In reply to Sasha Levin

From: Evan Quan <[email protected]>

[ Upstream commit 9822ba2ead1baa3de4860ad9472f652c4cc78c9c ]

Fix the compile error below:
drivers/gpu/drm/amd/amdgpu/../powerplay/smu_v11_0.c: In function 'smu_v11_0_init_microcode':

Quoted 1 line
Quoted 1 line

arch/arc/include/asm/bug.h:22:2: error: implicit declaration of function 'pr_warn'; did you mean 'pci_warn'? [-Werror=implicit-function-declaration]

      22 |  pr_warn("BUG: failure at %s:%d/%s()!\n", __FILE__, __LINE__, __func__); \
         |  ^~~~~~~
drivers/gpu/drm/amd/amdgpu/../powerplay/smu_v11_0.c:176:3: note: in expansion of macro 'BUG'
     176 |   BUG();
---
 drivers/gpu/drm/amd/powerplay/smu_v11_0.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
drivers/gpu/drm/amd/powerplay/smu_v11_0.c +2 -1
index 655ba4fb05dcd..48af305d42d54 100644
@@ -159,7 +159,8 @@ int smu_v11_0_init_microcode(struct smu_context *smu)
 		chip_name = "navi12";
 		break;
 	default:
-		BUG();
+		dev_err(adev->dev, "Unsupported ASIC type %d\n", adev->asic_type);
+		return -EINVAL;
 	}
 
 	snprintf(fw_name, sizeof(fw_name), "amdgpu/%s_smc.bin", chip_name);
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#37 lore raw

PATCH 36/60 bpf: Fix fds_example SIGSEGV error

In reply to Sasha Levin

From: Wenbo Zhang <[email protected]>

[ Upstream commit eef8a42d6ce087d1c81c960ae0d14f955b742feb ]

The BPF_LOG_BUF_SIZE's value is UINT32_MAX >> 8, so define an array with it on stack caused an overflow.

---
 samples/bpf/fds_example.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)
samples/bpf/fds_example.c +2 -1
index d5992f7872328..59f45fef51109 100644
@@ -30,6 +30,8 @@
 #define BPF_M_MAP	1
 #define BPF_M_PROG	2
 
+char bpf_log_buf[BPF_LOG_BUF_SIZE];
+
 static void usage(void)
 {
 	printf("Usage: fds_example [...]\n");
@@ -57,7 +59,6 @@ static int bpf_prog_create(const char *object)
 		BPF_EXIT_INSN(),
 	};
 	size_t insns_cnt = sizeof(insns) / sizeof(struct bpf_insn);
-	char bpf_log_buf[BPF_LOG_BUF_SIZE];
 	struct bpf_object *obj;
 	int prog_fd;
 
Signature · 1 line
2.25.1
author maintainer
#38 lore raw

PATCH 37/60 Bluetooth: hci_qca: Bug fixes for SSR

In reply to Sasha Levin

From: Venkata Lakshmi Narayana Gubba <[email protected]>

[ Upstream commit 3344537f614b966f726c1ec044d1c70a8cabe178 ]

1.During SSR for command time out if BT SoC goes to inresponsive state, power cycling of BT SoC was not happening. Given the fix by sending hw error event to reset the BT SoC.

2.If SSR is triggered then ignore the transmit data requests to BT SoC until SSR is completed.

---
 drivers/bluetooth/hci_qca.c | 40 +++++++++++++++++++++++++++++++++----
 1 file changed, 36 insertions(+), 4 deletions(-)
drivers/bluetooth/hci_qca.c +36 -4
index 0b1036e5e963c..14e4d2eaf8959 100644
@@ -71,7 +71,8 @@ enum qca_flags {
 	QCA_DROP_VENDOR_EVENT,
 	QCA_SUSPENDING,
 	QCA_MEMDUMP_COLLECTION,
-	QCA_HW_ERROR_EVENT
+	QCA_HW_ERROR_EVENT,
+	QCA_SSR_TRIGGERED
 };
 
 
@@ -854,6 +855,13 @@ static int qca_enqueue(struct hci_uart *hu, struct sk_buff *skb)
 	BT_DBG("hu %p qca enq skb %p tx_ibs_state %d", hu, skb,
 	       qca->tx_ibs_state);
 
+	if (test_bit(QCA_SSR_TRIGGERED, &qca->flags)) {
+		/* As SSR is in progress, ignore the packets */
+		bt_dev_dbg(hu->hdev, "SSR is in progress");
+		kfree_skb(skb);
+		return 0;
+	}
+
 	/* Prepend skb with frame type */
 	memcpy(skb_push(skb, 1), &hci_skb_pkt_type(skb), 1);
 
@@ -1085,6 +1093,7 @@ static int qca_controller_memdump_event(struct hci_dev *hdev,
 	struct hci_uart *hu = hci_get_drvdata(hdev);
 	struct qca_data *qca = hu->priv;
 
+	set_bit(QCA_SSR_TRIGGERED, &qca->flags);
 	skb_queue_tail(&qca->rx_memdump_q, skb);
 	queue_work(qca->workqueue, &qca->ctrl_memdump_evt);
 
@@ -1445,6 +1454,7 @@ static void qca_hw_error(struct hci_dev *hdev, u8 code)
 	struct qca_memdump_data *qca_memdump = qca->qca_memdump;
 	char *memdump_buf = NULL;
 
+	set_bit(QCA_SSR_TRIGGERED, &qca->flags);
 	set_bit(QCA_HW_ERROR_EVENT, &qca->flags);
 	bt_dev_info(hdev, "mem_dump_status: %d", qca->memdump_state);
 
@@ -1489,10 +1499,30 @@ static void qca_cmd_timeout(struct hci_dev *hdev)
 	struct hci_uart *hu = hci_get_drvdata(hdev);
 	struct qca_data *qca = hu->priv;
 
-	if (qca->memdump_state == QCA_MEMDUMP_IDLE)
+	set_bit(QCA_SSR_TRIGGERED, &qca->flags);
+	if (qca->memdump_state == QCA_MEMDUMP_IDLE) {
+		set_bit(QCA_MEMDUMP_COLLECTION, &qca->flags);
 		qca_send_crashbuffer(hu);
-	else
-		bt_dev_info(hdev, "Dump collection is in process");
+		qca_wait_for_dump_collection(hdev);
+	} else if (qca->memdump_state == QCA_MEMDUMP_COLLECTING) {
+		/* Let us wait here until memory dump collected or
+		 * memory dump timer expired.
+		 */
+		bt_dev_info(hdev, "waiting for dump to complete");
+		qca_wait_for_dump_collection(hdev);
+	}
+
+	mutex_lock(&qca->hci_memdump_lock);
+	if (qca->memdump_state != QCA_MEMDUMP_COLLECTED) {
+		qca->memdump_state = QCA_MEMDUMP_TIMEOUT;
+		if (!test_bit(QCA_HW_ERROR_EVENT, &qca->flags)) {
+			/* Inject hw error event to reset the device
+			 * and driver.
+			 */
+			hci_reset_dev(hu->hdev);
+		}
+	}
+	mutex_unlock(&qca->hci_memdump_lock);
 }
 
 static int qca_wcn3990_init(struct hci_uart *hu)
@@ -1603,6 +1633,8 @@ static int qca_setup(struct hci_uart *hu)
 	if (ret)
 		return ret;
 
+	clear_bit(QCA_SSR_TRIGGERED, &qca->flags);
+
 	if (qca_is_wcn399x(soc_type)) {
 		set_bit(HCI_QUIRK_USE_BDADDR_PROPERTY, &hdev->quirks);
 
Signature · 1 line
2.25.1
author maintainer
#39 lore raw

PATCH 38/60 brcmfmac: keep SDIO watchdog running when console_interval is non-zero

In reply to Sasha Levin

From: Wright Feng <[email protected]>

[ Upstream commit eccbf46b15bb3e35d004148f7c3a8fa8e9b26c1e ]

brcmfmac host driver makes SDIO bus sleep and stops SDIO watchdog if no pending event or data. As a result, host driver does not poll firmware console buffer before buffer overflow, which leads to missing firmware logs. We should not stop SDIO watchdog if console_interval is non-zero in debug build.

---
 drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/sdio.c +5 -1
index 3a08252f1a53f..0dbbb467c2291 100644
@@ -3689,7 +3689,11 @@ static void brcmf_sdio_bus_watchdog(struct brcmf_sdio *bus)
 			if (bus->idlecount > bus->idletime) {
 				brcmf_dbg(SDIO, "idle\n");
 				sdio_claim_host(bus->sdiodev->func1);
-				brcmf_sdio_wd_timer(bus, false);
+#ifdef DEBUG
+				if (!BRCMF_FWCON_ON() ||
+				    bus->console_interval == 0)
+#endif
+					brcmf_sdio_wd_timer(bus, false);
 				bus->idlecount = 0;
 				brcmf_sdio_bus_sleep(bus, true, false);
 				sdio_release_host(bus->sdiodev->func1);
Signature · 1 line
2.25.1
author maintainer
#40 lore raw

PATCH 39/60 brcmfmac: To fix Bss Info flag definition Bug

In reply to Sasha Levin

From: Prasanna Kerekoppa <[email protected]>

[ Upstream commit fa3266541b13f390eb35bdbc38ff4a03368be004 ]

Bss info flag definition need to be fixed from 0x2 to 0x4 This flag is for rssi info received on channel.
All Firmware branches defined as 0x4 and this is bug in brcmfmac.

---
 drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwil_types.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwil_types.h +1 -1
index de0ef1b545c4f..2e31cc10c1954 100644
@@ -19,7 +19,7 @@
 #define BRCMF_ARP_OL_PEER_AUTO_REPLY	0x00000008
 
 #define	BRCMF_BSS_INFO_VERSION	109 /* curr ver of brcmf_bss_info_le struct */
-#define BRCMF_BSS_RSSI_ON_CHANNEL	0x0002
+#define BRCMF_BSS_RSSI_ON_CHANNEL	0x0004
 
 #define BRCMF_STA_BRCM			0x00000001	/* Running a Broadcom driver */
 #define BRCMF_STA_WME			0x00000002	/* WMM association */
Signature · 1 line
2.25.1
author maintainer
#41 lore raw

PATCH 40/60 brcmfmac: set state of hanger slot to FREE when flushing PSQ

In reply to Sasha Levin

From: Wright Feng <[email protected]>

[ Upstream commit fcdd7a875def793c38d7369633af3eba6c7cf089 ]

When USB or SDIO device got abnormal bus disconnection, host driver tried to clean up the skbs in PSQ and TXQ (The skb's pointer in hanger slot linked to PSQ and TSQ), so we should set the state of skb hanger slot to BRCMF_FWS_HANGER_ITEM_STATE_FREE before freeing skb.
In brcmf_fws_bus_txq_cleanup it already sets BRCMF_FWS_HANGER_ITEM_STATE_FREE before freeing skb, therefore we add the same thing in brcmf_fws_psq_flush to avoid following warning message.

   [ 1580.012880] ------------   [ cut here ]------------
   [ 1580.017550] WARNING: CPU: 3 PID: 3065 at
drivers/net/wireless/broadcom/brcm80211/brcmutil/utils.c:49
brcmu_pkt_buf_free_skb+0x21/0x30 [brcmutil]
   [ 1580.184017] Call Trace:
   [ 1580.186514]  brcmf_fws_cleanup+0x14e/0x190 [brcmfmac]
   [ 1580.191594]  brcmf_fws_del_interface+0x70/0x90 [brcmfmac]
   [ 1580.197029]  brcmf_proto_bcdc_del_if+0xe/0x10 [brcmfmac]
   [ 1580.202418]  brcmf_remove_interface+0x69/0x190 [brcmfmac]
   [ 1580.207888]  brcmf_detach+0x90/0xe0 [brcmfmac]
   [ 1580.212385]  brcmf_usb_disconnect+0x76/0xb0 [brcmfmac]
   [ 1580.217557]  usb_unbind_interface+0x72/0x260
   [ 1580.221857]  device_release_driver_internal+0x141/0x200
   [ 1580.227152]  device_release_driver+0x12/0x20
   [ 1580.231460]  bus_remove_device+0xfd/0x170
   [ 1580.235504]  device_del+0x1d9/0x300
   [ 1580.239041]  usb_disable_device+0x9e/0x270
   [ 1580.243160]  usb_disconnect+0x94/0x270
   [ 1580.246980]  hub_event+0x76d/0x13b0
   [ 1580.250499]  process_one_work+0x144/0x360
   [ 1580.254564]  worker_thread+0x4d/0x3c0
   [ 1580.258247]  kthread+0x109/0x140
   [ 1580.261515]  ? rescuer_thread+0x340/0x340
   [ 1580.265543]  ? kthread_park+0x60/0x60
   [ 1580.269237]  ? SyS_exit_group+0x14/0x20
   [ 1580.273118]  ret_from_fork+0x25/0x30
   [ 1580.300446] ------------   [ cut here ]------------
---
 drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwsignal.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/net/wireless/broadcom/brcm80211/brcmfmac/fwsignal.c +4 -0
index 8cc52935fd413..948840b4e38e3 100644
@@ -643,6 +643,7 @@ static inline int brcmf_fws_hanger_poppkt(struct brcmf_fws_hanger *h,
 static void brcmf_fws_psq_flush(struct brcmf_fws_info *fws, struct pktq *q,
 				int ifidx)
 {
+	struct brcmf_fws_hanger_item *hi;
 	bool (*matchfn)(struct sk_buff *, void *) = NULL;
 	struct sk_buff *skb;
 	int prec;
@@ -654,6 +655,9 @@ static void brcmf_fws_psq_flush(struct brcmf_fws_info *fws, struct pktq *q,
 		skb = brcmu_pktq_pdeq_match(q, prec, matchfn, &ifidx);
 		while (skb) {
 			hslot = brcmf_skb_htod_tag_get_field(skb, HSLOT);
+			hi = &fws->hanger.items[hslot];
+			WARN_ON(skb != hi->pkt);
+			hi->state = BRCMF_FWS_HANGER_ITEM_STATE_FREE;
 			brcmf_fws_hanger_poppkt(&fws->hanger, hslot, &skb,
 						true);
 			brcmu_pkt_buf_free_skb(skb);
Signature · 1 line
2.25.1
author maintainer
#42 lore raw

PATCH 41/60 platform/x86: asus-nb-wmi: add support for ASUS ROG Zephyrus G14 and G15

In reply to Sasha Levin

From: Armas Spann <[email protected]>

[ Upstream commit 13bceda68fb9ef388ad40d355ab8d03ee64d14c2 ]

Add device support for the new ASUS ROG Zephyrus G14 (GA401I) and G15 (GA502I) series.

This is accomplished by two new quirk entries (one per each series), as well as all current available G401I/G502I DMI_PRODUCT_NAMEs to match the corresponding devices.

---
 drivers/platform/x86/asus-nb-wmi.c | 82 ++++++++++++++++++++++++++++++
 1 file changed, 82 insertions(+)
drivers/platform/x86/asus-nb-wmi.c +82 -0
index c4404d9c1de4f..1bb082308c209 100644
@@ -110,6 +110,16 @@ static struct quirk_entry quirk_asus_forceals = {
 	.wmi_force_als_set = true,
 };
 
+static struct quirk_entry quirk_asus_ga401i = {
+	.wmi_backlight_power = true,
+	.wmi_backlight_set_devstate = true,
+};
+
+static struct quirk_entry quirk_asus_ga502i = {
+	.wmi_backlight_power = true,
+	.wmi_backlight_set_devstate = true,
+};
+
 static int dmi_matched(const struct dmi_system_id *dmi)
 {
 	pr_info("Identified laptop model '%s'\n", dmi->ident);
@@ -411,6 +421,78 @@ static const struct dmi_system_id asus_quirks[] = {
 		},
 		.driver_data = &quirk_asus_forceals,
 	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA401IH",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IH"),
+		},
+		.driver_data = &quirk_asus_ga401i,
+	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA401II",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA401II"),
+		},
+		.driver_data = &quirk_asus_ga401i,
+	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA401IU",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IU"),
+		},
+		.driver_data = &quirk_asus_ga401i,
+	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA401IV",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IV"),
+		},
+		.driver_data = &quirk_asus_ga401i,
+	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA401IVC",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA401IVC"),
+		},
+		.driver_data = &quirk_asus_ga401i,
+	},
+		{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA502II",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA502II"),
+		},
+		.driver_data = &quirk_asus_ga502i,
+	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA502IU",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA502IU"),
+		},
+		.driver_data = &quirk_asus_ga502i,
+	},
+	{
+		.callback = dmi_matched,
+		.ident = "ASUSTeK COMPUTER INC. GA502IV",
+		.matches = {
+			DMI_MATCH(DMI_SYS_VENDOR, "ASUSTeK COMPUTER INC."),
+			DMI_MATCH(DMI_PRODUCT_NAME, "GA502IV"),
+		},
+		.driver_data = &quirk_asus_ga502i,
+	},
 	{},
 };
 
Signature · 1 line
2.25.1
author maintainer
#43 lore raw

PATCH 42/60 iwlegacy: Check the return value of pcie_capability_read_*()

In reply to Sasha Levin

From: Bolarinwa Olayemi Saheed <[email protected]>

[ Upstream commit 9018fd7f2a73e9b290f48a56b421558fa31e8b75 ]

On failure pcie_capability_read_dword() sets it's last parameter, val to 0. However, with Patch 14/14, it is possible that val is set to ~0 on failure. This would introduce a bug because (x & x) == (~0 & x).

This bug can be avoided without changing the function's behaviour if the return value of pcie_capability_read_dword is checked to confirm success.

Check the return value of pcie_capability_read_dword() to ensure success.

---
 drivers/net/wireless/intel/iwlegacy/common.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/net/wireless/intel/iwlegacy/common.c +2 -2
index 348c17ce72f5c..f78e062df572a 100644
@@ -4286,8 +4286,8 @@ il_apm_init(struct il_priv *il)
 	 *    power savings, even without L1.
 	 */
 	if (il->cfg->set_l0s) {
-		pcie_capability_read_word(il->pci_dev, PCI_EXP_LNKCTL, &lctl);
-		if (lctl & PCI_EXP_LNKCTL_ASPM_L1) {
+		ret = pcie_capability_read_word(il->pci_dev, PCI_EXP_LNKCTL, &lctl);
+		if (!ret && (lctl & PCI_EXP_LNKCTL_ASPM_L1)) {
 			/* L1-ASPM enabled; disable(!) L0S  */
 			il_set_bit(il, CSR_GIO_REG,
 				   CSR_GIO_REG_VAL_L0S_ENABLED);
Signature · 1 line
2.25.1
author maintainer
#44 lore raw

PATCH 43/60 gpu: host1x: debug: Fix multiple channels emitting messages simultaneously

In reply to Sasha Levin

From: Dmitry Osipenko <[email protected]>

[ Upstream commit 35681862808472a0a4b9a8817ae2789c0b5b3edc ]

Once channel's job is hung, it dumps the channel's state into KMSG before tearing down the offending job. If multiple channels hang at once, then they dump messages simultaneously, making the debug info unreadable, and thus, useless. This patch adds mutex which allows only one channel to emit debug messages at a time.

---
 drivers/gpu/host1x/debug.c | 4 ++++
 1 file changed, 4 insertions(+)
drivers/gpu/host1x/debug.c +4 -0
index c0392672a8421..1b4997bda1c79 100644
@@ -16,6 +16,8 @@
 #include "debug.h"
 #include "channel.h"
 
+static DEFINE_MUTEX(debug_lock);
+
 unsigned int host1x_debug_trace_cmdbuf;
 
 static pid_t host1x_debug_force_timeout_pid;
@@ -52,12 +54,14 @@ static int show_channel(struct host1x_channel *ch, void *data, bool show_fifo)
 	struct output *o = data;
 
 	mutex_lock(&ch->cdma.lock);
+	mutex_lock(&debug_lock);
 
 	if (show_fifo)
 		host1x_hw_show_channel_fifo(m, ch, o);
 
 	host1x_hw_show_channel_cdma(m, ch, o);
 
+	mutex_unlock(&debug_lock);
 	mutex_unlock(&ch->cdma.lock);
 
 	return 0;
Signature · 6 lines
2.25.1

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

PATCH 44/60 drm/amd/powerplay: suppress compile error around BUG_ON

In reply to Sasha Levin

From: Evan Quan <[email protected]>

[ Upstream commit 75bc07e2403caea9ecac69f766dfb7dc33547594 ]

To suppress the compile error below for "ARCH=arc".
drivers/gpu/drm/amd/amdgpu/../powerplay/arcturus_ppt.c: In function 'arcturus_fill_eeprom_i2c_req':

Quoted 1 line
Quoted 1 line

arch/arc/include/asm/bug.h:22:2: error: implicit declaration of function 'pr_warn'; did you mean 'pci_warn'? [-Werror=implicit-function-declaration]

      22 |  pr_warn("BUG: failure at %s:%d/%s()!\n", __FILE__, __LINE__, __func__); \
         |  ^~~~~~~
   include/asm-generic/bug.h:62:57: note: in expansion of macro 'BUG'
      62 | #define BUG_ON(condition) do { if (unlikely(condition)) BUG(); } while (0)
         |                                                         ^~~
   drivers/gpu/drm/amd/amdgpu/../powerplay/arcturus_ppt.c:2157:2: note: in expansion of macro 'BUG_ON'
    2157 |  BUG_ON(numbytes > MAX_SW_I2C_COMMANDS);
---
 drivers/gpu/drm/amd/powerplay/arcturus_ppt.c | 14 ++++++++++++--
 1 file changed, 12 insertions(+), 2 deletions(-)
drivers/gpu/drm/amd/powerplay/arcturus_ppt.c +12 -2
index 1ef0923f71906..9ad0e6f18be49 100644
@@ -2035,8 +2035,6 @@ static void arcturus_fill_eeprom_i2c_req(SwI2cRequest_t  *req, bool write,
 {
 	int i;
 
-	BUG_ON(numbytes > MAX_SW_I2C_COMMANDS);
-
 	req->I2CcontrollerPort = 0;
 	req->I2CSpeed = 2;
 	req->SlaveAddress = address;
@@ -2074,6 +2072,12 @@ static int arcturus_i2c_eeprom_read_data(struct i2c_adapter *control,
 	struct smu_table_context *smu_table = &adev->smu.smu_table;
 	struct smu_table *table = &smu_table->driver_table;
 
+	if (numbytes > MAX_SW_I2C_COMMANDS) {
+		dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n",
+			numbytes, MAX_SW_I2C_COMMANDS);
+		return -EINVAL;
+	}
+
 	memset(&req, 0, sizeof(req));
 	arcturus_fill_eeprom_i2c_req(&req, false, address, numbytes, data);
 
@@ -2110,6 +2114,12 @@ static int arcturus_i2c_eeprom_write_data(struct i2c_adapter *control,
 	SwI2cRequest_t req;
 	struct amdgpu_device *adev = to_amdgpu_device(control);
 
+	if (numbytes > MAX_SW_I2C_COMMANDS) {
+		dev_err(adev->dev, "numbytes requested %d is over max allowed %d\n",
+			numbytes, MAX_SW_I2C_COMMANDS);
+		return -EINVAL;
+	}
+
 	memset(&req, 0, sizeof(req));
 	arcturus_fill_eeprom_i2c_req(&req, true, address, numbytes, data);
 
Signature · 6 lines
2.25.1

_______________________________________________
amd-gfx mailing list
[email protected]
https://lists.freedesktop.org/mailman/listinfo/amd-gfx
author maintainer
#46 lore raw

PATCH 45/60 ionic: update eid test for overflow

In reply to Sasha Levin

From: Shannon Nelson <[email protected]>

[ Upstream commit 3fbc9bb6ca32d12d4d32a7ae32abef67ac95f889 ]

Fix up our comparison to better handle a potential (but largely unlikely) wrap around.

---
 drivers/net/ethernet/pensando/ionic/ionic_lif.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/net/ethernet/pensando/ionic/ionic_lif.c +1 -1
index 337d971ffd92c..29f77faa808bb 100644
@@ -709,7 +709,7 @@ static bool ionic_notifyq_service(struct ionic_cq *cq,
 	eid = le64_to_cpu(comp->event.eid);
 
 	/* Have we run out of new completions to process? */
-	if (eid <= lif->last_eid)
+	if ((s64)(eid - lif->last_eid) <= 0)
 		return false;
 
 	lif->last_eid = eid;
Signature · 1 line
2.25.1
author maintainer
#47 lore raw

PATCH 46/60 x86/uaccess: Make __get_user_size() Clang compliant on 32-bit

In reply to Sasha Levin

From: Nick Desaulniers <[email protected]>

[ Upstream commit 158807de5822d1079e162a3762956fd743dd483e ]

Clang fails to compile __get_user_size() on 32-bit for the following code:

      long long val;

      __get_user(val, usrptr);

with: error: invalid output size for constraint '=q'

GCC compiles the same code without complaints.

The reason is that GCC and Clang are architecturally different, which leads to subtle issues for code that's invalid but clearly dead, i.e. with code that emulates polymorphism with the preprocessor and sizeof.

GCC will perform semantic analysis after early inlining and dead code elimination, so it will not warn on invalid code that's dead. Clang strictly performs optimizations after semantic analysis, so it will warn for dead code.

Neither Clang nor GCC like this very much with -m32:

long long ret;
asm ("movb $5, %0" : "=q" (ret));

However, GCC can tolerate this variant:

long long ret;
switch (sizeof(ret)) {
case 1:

        asm ("movb $5, %0" : "=q" (ret));
        break;

case 8:;
}

Clang, on the other hand, won't accept that because it validates the inline asm for the '1' case before the optimisation phase where it realises that it wouldn't have to emit it anyway.

If LLVM (Clang's "back end") fails such as during instruction selection or register allocation, it cannot provide accurate diagnostics (warnings / errors) that contain line information, as the AST has been discarded from memory at that point.

While there have been early discussions about having C/C++ specific language optimizations in Clang via the use of MLIR, which would enable such earlier optimizations, such work is not scoped and likely a multi-year endeavor.

It was discussed to change the asm output constraint for the one byte case from "=q" to "=r". While it works for 64-bit, it fails on 32-bit. With '=r' the compiler could fail to chose a register accessible as high/low which is required for the byte operation. If that happens the assembly will fail.

Use a local temporary variable of type 'unsigned char' as output for the byte copy inline asm and then assign it to the real output variable. This prevents Clang from failing the semantic analysis in the above case.

The resulting code for the actual one byte copy is not affected as the temporary variable is optimized out.

[ tglx: Amended changelog ]

---
 arch/x86/include/asm/uaccess.h | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)
arch/x86/include/asm/uaccess.h +4 -1
index d8f283b9a569c..d1323c73cf6d2 100644
@@ -314,11 +314,14 @@ do {									\
 
 #define __get_user_size(x, ptr, size, retval)				\
 do {									\
+	unsigned char x_u8__;						\
+									\
 	retval = 0;							\
 	__chk_user_ptr(ptr);						\
 	switch (size) {							\
 	case 1:								\
-		__get_user_asm(x, ptr, retval, "b", "=q");		\
+		__get_user_asm(x_u8__, ptr, retval, "b", "=q");		\
+		(x) = x_u8__;						\
 		break;							\
 	case 2:								\
 		__get_user_asm(x, ptr, retval, "w", "=r");		\
Signature · 1 line
2.25.1
author maintainer
#48 lore raw

PATCH 47/60 mmc: sdhci-pci-o2micro: Bug fix for O2 host controller Seabird1

In reply to Sasha Levin

From: shirley her <[email protected]>

[ Upstream commit cdd2b769789ae1a030e1a26f6c37c5833cabcb34 ]

To fix support for the O2 host controller Seabird1, set the quirk SDHCI_QUIRK2_PRESET_VALUE_BROKEN and the capability bit MMC_CAP2_NO_SDIO. Moreover, assign the ->get_cd() callback.

---
 drivers/mmc/host/sdhci-pci-o2micro.c | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/mmc/host/sdhci-pci-o2micro.c +6 -0
index fa8105087d684..41a2394313dd0 100644
@@ -561,6 +561,12 @@ int sdhci_pci_o2_probe_slot(struct sdhci_pci_slot *slot)
 			slot->host->mmc_host_ops.get_cd = sdhci_o2_get_cd;
 		}
 
+		if (chip->pdev->device == PCI_DEVICE_ID_O2_SEABIRD1) {
+			slot->host->mmc_host_ops.get_cd = sdhci_o2_get_cd;
+			host->mmc->caps2 |= MMC_CAP2_NO_SDIO;
+			host->quirks2 |= SDHCI_QUIRK2_PRESET_VALUE_BROKEN;
+		}
+
 		host->mmc_host_ops.execute_tuning = sdhci_o2_execute_tuning;
 
 		if (chip->pdev->device != PCI_DEVICE_ID_O2_FUJIN2)
Signature · 1 line
2.25.1
author maintainer
#49 lore raw

PATCH 48/60 usb: gadget: net2280: fix memory leak on probe error handling paths

In reply to Sasha Levin

From: Evgeny Novikov <[email protected]>

[ Upstream commit 2468c877da428ebfd701142c4cdfefcfb7d4c00e ]

Driver does not release memory for device on error handling paths in net2280_probe() when gadget_release() is not registered yet.

The patch fixes the bug like in other similar drivers.

Found by Linux Driver Verification project (linuxtesting.org).

---
 drivers/usb/gadget/udc/net2280.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/usb/gadget/udc/net2280.c +3 -1
index 5eff85eeaa5a0..7530bd9a08c43 100644
@@ -3781,8 +3781,10 @@ static int net2280_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 	return 0;
 
 done:
-	if (dev)
+	if (dev) {
 		net2280_remove(pdev);
+		kfree(dev);
+	}
 	return retval;
 }
 
Signature · 1 line
2.25.1
author maintainer
#50 lore raw

PATCH 49/60 bdc: Fix bug causing crash after multiple disconnects

In reply to Sasha Levin

From: Sasi Kumar <[email protected]>

[ Upstream commit a95bdfd22076497288868c028619bc5995f5cc7f ]

Multiple connects/disconnects can cause a crash on the second disconnect. The driver had a problem where it would try to send endpoint commands after it was disconnected which is not allowed by the hardware. The fix is to only allow the endpoint commands when the endpoint is connected. This will also fix issues that showed up when using configfs to create gadgets.

---
 drivers/usb/gadget/udc/bdc/bdc_core.c |  4 ++++
 drivers/usb/gadget/udc/bdc/bdc_ep.c   | 16 ++++++++++------
 2 files changed, 14 insertions(+), 6 deletions(-)
drivers/usb/gadget/udc/bdc/bdc_core.c +4 -0
index 02a3a774670b1..5fde5a8b065c1 100644
@@ -282,6 +282,7 @@ static void bdc_mem_init(struct bdc *bdc, bool reinit)
 	 * in that case reinit is passed as 1
 	 */
 	if (reinit) {
+		int i;
 		/* Enable interrupts */
 		temp = bdc_readl(bdc->regs, BDC_BDCSC);
 		temp |= BDC_GIE;
@@ -291,6 +292,9 @@ static void bdc_mem_init(struct bdc *bdc, bool reinit)
 		/* Initialize SRR to 0 */
 		memset(bdc->srr.sr_bds, 0,
 					NUM_SR_ENTRIES * sizeof(struct bdc_bd));
+		/* clear ep flags to avoid post disconnect stops/deconfigs */
+		for (i = 1; i < bdc->num_eps; ++i)
+			bdc->bdc_ep_array[i]->flags = 0;
 	} else {
 		/* One time initiaization only */
 		/* Enable status report function pointers */
drivers/usb/gadget/udc/bdc/bdc_ep.c +10 -6
index d49c6dc1082dc..9ddc0b4e92c9c 100644
@@ -615,7 +615,6 @@ int bdc_ep_enable(struct bdc_ep *ep)
 	}
 	bdc_dbg_bd_list(bdc, ep);
 	/* only for ep0: config ep is called for ep0 from connect event */
-	ep->flags |= BDC_EP_ENABLED;
 	if (ep->ep_num == 1)
 		return ret;
 
@@ -759,10 +758,13 @@ static int ep_dequeue(struct bdc_ep *ep, struct bdc_req *req)
 					__func__, ep->name, start_bdi, end_bdi);
 	dev_dbg(bdc->dev, "ep_dequeue ep=%p ep->desc=%p\n",
 						ep, (void *)ep->usb_ep.desc);
-	/* Stop the ep to see where the HW is ? */
-	ret = bdc_stop_ep(bdc, ep->ep_num);
-	/* if there is an issue with stopping ep, then no need to go further */
-	if (ret)
+	/* if still connected, stop the ep to see where the HW is ? */
+	if (!(bdc_readl(bdc->regs, BDC_USPC) & BDC_PST_MASK)) {
+		ret = bdc_stop_ep(bdc, ep->ep_num);
+		/* if there is an issue, then no need to go further */
+		if (ret)
+			return 0;
+	} else
 		return 0;
 
 	/*
@@ -1911,7 +1913,9 @@ static int bdc_gadget_ep_disable(struct usb_ep *_ep)
 		__func__, ep->name, ep->flags);
 
 	if (!(ep->flags & BDC_EP_ENABLED)) {
-		dev_warn(bdc->dev, "%s is already disabled\n", ep->name);
+		if (bdc->gadget.speed != USB_SPEED_UNKNOWN)
+			dev_warn(bdc->dev, "%s is already disabled\n",
+				 ep->name);
 		return 0;
 	}
 	spin_lock_irqsave(&bdc->lock, flags);
Signature · 1 line
2.25.1
author maintainer
#51 lore raw

PATCH 50/60 usb: bdc: Halt controller on suspend

In reply to Sasha Levin

From: Danesh Petigara <[email protected]>

[ Upstream commit 5fc453d7de3d0c345812453823a3a56783c5f82c ]

GISB bus error kernel panics have been observed during S2 transition tests on the 7271t platform. The errors are a result of the BDC interrupt handler trying to access BDC register space after the system's suspend callbacks have completed.

Adding a suspend hook to the BDC driver that halts the controller before S2 entry thus preventing unwanted access to the BDC register space during this transition.

---
 drivers/usb/gadget/udc/bdc/bdc_core.c | 9 +++++++--
 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/usb/gadget/udc/bdc/bdc_core.c +7 -2
index 5fde5a8b065c1..2dca11f0a7444 100644
@@ -603,9 +603,14 @@ static int bdc_remove(struct platform_device *pdev)
 static int bdc_suspend(struct device *dev)
 {
 	struct bdc *bdc = dev_get_drvdata(dev);
+	int ret;
 
-	clk_disable_unprepare(bdc->clk);
-	return 0;
+	/* Halt the controller */
+	ret = bdc_stop(bdc);
+	if (!ret)
+		clk_disable_unprepare(bdc->clk);
+
+	return ret;
 }
 
 static int bdc_resume(struct device *dev)
Signature · 1 line
2.25.1
author maintainer
#52 lore raw

PATCH 51/60 dyndbg: fix a BUG_ON in ddebug_describe_flags

In reply to Sasha Levin

From: Jim Cromie <[email protected]>

[ Upstream commit f678ce8cc3cb2ad29df75d8824c74f36398ba871 ]

ddebug_describe_flags() currently fills a caller provided string buffer, after testing its size (also passed) in a BUG_ON. Fix this by replacing them with a known-big-enough string buffer wrapped in a struct, and passing that instead.

Also simplify ddebug_describe_flags() flags parameter from a struct to a member in that struct, and hoist the member deref up to the caller. This makes the function reusable (soon) where flags are unpacked.

---
 lib/dynamic_debug.c | 23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)
lib/dynamic_debug.c +11 -12
index 8f199f403ab50..e3755d1f74bd2 100644
@@ -87,22 +87,22 @@ static struct { unsigned flag:8; char opt_char; } opt_array[] = {
 	{ _DPRINTK_FLAGS_NONE, '_' },
 };
 
+struct flagsbuf { char buf[ARRAY_SIZE(opt_array)+1]; };
+
 /* format a string into buf[] which describes the _ddebug's flags */
-static char *ddebug_describe_flags(struct _ddebug *dp, char *buf,
-				    size_t maxlen)
+static char *ddebug_describe_flags(unsigned int flags, struct flagsbuf *fb)
 {
-	char *p = buf;
+	char *p = fb->buf;
 	int i;
 
-	BUG_ON(maxlen < 6);
 	for (i = 0; i < ARRAY_SIZE(opt_array); ++i)
-		if (dp->flags & opt_array[i].flag)
+		if (flags & opt_array[i].flag)
 			*p++ = opt_array[i].opt_char;
-	if (p == buf)
+	if (p == fb->buf)
 		*p++ = '_';
 	*p = '\0';
 
-	return buf;
+	return fb->buf;
 }
 
 #define vpr_info(fmt, ...)					\
@@ -144,7 +144,7 @@ static int ddebug_change(const struct ddebug_query *query,
 	struct ddebug_table *dt;
 	unsigned int newflags;
 	unsigned int nfound = 0;
-	char flagbuf[10];
+	struct flagsbuf fbuf;
 
 	/* search for matching ddebugs */
 	mutex_lock(&ddebug_lock);
@@ -201,8 +201,7 @@ static int ddebug_change(const struct ddebug_query *query,
 			vpr_info("changed %s:%d [%s]%s =%s\n",
 				 trim_prefix(dp->filename), dp->lineno,
 				 dt->mod_name, dp->function,
-				 ddebug_describe_flags(dp, flagbuf,
-						       sizeof(flagbuf)));
+				 ddebug_describe_flags(dp->flags, &fbuf));
 		}
 	}
 	mutex_unlock(&ddebug_lock);
@@ -816,7 +815,7 @@ static int ddebug_proc_show(struct seq_file *m, void *p)
 {
 	struct ddebug_iter *iter = m->private;
 	struct _ddebug *dp = p;
-	char flagsbuf[10];
+	struct flagsbuf flags;
 
 	vpr_info("called m=%p p=%p\n", m, p);
 
@@ -829,7 +828,7 @@ static int ddebug_proc_show(struct seq_file *m, void *p)
 	seq_printf(m, "%s:%u [%s]%s =%s \"",
 		   trim_prefix(dp->filename), dp->lineno,
 		   iter->table->mod_name, dp->function,
-		   ddebug_describe_flags(dp, flagsbuf, sizeof(flagsbuf)));
+		   ddebug_describe_flags(dp->flags, &flags));
 	seq_escape(m, dp->format, "\t\r\n\"");
 	seq_puts(m, "\"\n");
 
Signature · 1 line
2.25.1
author maintainer
#53 lore raw

PATCH 52/60 dyndbg: prefer declarative init in caller, to memset in callee

In reply to Sasha Levin

From: Jim Cromie <[email protected]>

[ Upstream commit 9c9d0acbe2793315fa6945a19685ad2a51fb281b ]

ddebug_exec_query declares an auto var, and passes it to ddebug_parse_query, which memsets it before using it. Drop that memset, instead initialize the variable in the caller; let the compiler decide how to do it.

---
 lib/dynamic_debug.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
lib/dynamic_debug.c +1 -2
index e3755d1f74bd2..4f0bd560478f7 100644
@@ -327,7 +327,6 @@ static int ddebug_parse_query(char *words[], int nwords,
 		pr_err("expecting pairs of match-spec <value>\n");
 		return -EINVAL;
 	}
-	memset(query, 0, sizeof(*query));
 
 	if (modname)
 		/* support $modname.dyndbg=<multiple queries> */
@@ -445,7 +444,7 @@ static int ddebug_parse_flags(const char *str, unsigned int *flagsp,
 static int ddebug_exec_query(char *query_string, const char *modname)
 {
 	unsigned int flags = 0, mask = 0;
-	struct ddebug_query query;
+	struct ddebug_query query = {};
 #define MAXWORDS 9
 	int nwords, nfound;
 	char *words[MAXWORDS];
Signature · 1 line
2.25.1
#54 lore raw

Re: [PATCH AUTOSEL 5.7 52/60] dyndbg: prefer declarative init in caller, to memset in callee

Quoted 42 lines · Sasha Levin

From: Jim Cromie <[email protected]>

[ Upstream commit 9c9d0acbe2793315fa6945a19685ad2a51fb281b ]

ddebug_exec_query declares an auto var, and passes it to ddebug_parse_query, which memsets it before using it. Drop that memset, instead initialize the variable in the caller; let the compiler decide how to do it.

---
 lib/dynamic_debug.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
lib/dynamic_debug.c +1 -2
index e3755d1f74bd2..4f0bd560478f7 100644
@@ -327,7 +327,6 @@ static int ddebug_parse_query(char *words[], int nwords,
 		pr_err("expecting pairs of match-spec <value>\n");
 		return -EINVAL;
 	}
-	memset(query, 0, sizeof(*query));
 
 	if (modname)
 		/* support $modname.dyndbg=<multiple queries> */
@@ -445,7 +444,7 @@ static int ddebug_parse_flags(const char *str, unsigned int *flagsp,
 static int ddebug_exec_query(char *query_string, const char *modname)
 {
 	unsigned int flags = 0, mask = 0;
-	struct ddebug_query query;
+	struct ddebug_query query = {};
 #define MAXWORDS 9
 	int nwords, nfound;
 	char *words[MAXWORDS];
Signature · 1 line
2.25.1

There's no need for this in stable kernels, please drop it from everywhere.

thanks,

greg k-h

author maintainer
#55 lore raw

Re: [PATCH AUTOSEL 5.7 52/60] dyndbg: prefer declarative init in caller, to memset in callee

Quoted 38 lines · Greg Kroah-Hartman
Quoted 37 lines · Sasha Levin

From: Jim Cromie <[email protected]>

[ Upstream commit 9c9d0acbe2793315fa6945a19685ad2a51fb281b ]

ddebug_exec_query declares an auto var, and passes it to ddebug_parse_query, which memsets it before using it. Drop that memset, instead initialize the variable in the caller; let the compiler decide how to do it.

---
 lib/dynamic_debug.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
lib/dynamic_debug.c +1 -2
index e3755d1f74bd2..4f0bd560478f7 100644
@@ -327,7 +327,6 @@ static int ddebug_parse_query(char *words[], int nwords,
 		pr_err("expecting pairs of match-spec <value>\n");
 		return -EINVAL;
 	}
-	memset(query, 0, sizeof(*query));
 
 	if (modname)
 		/* support $modname.dyndbg=<multiple queries> */
@@ -445,7 +444,7 @@ static int ddebug_parse_flags(const char *str, unsigned int *flagsp,
 static int ddebug_exec_query(char *query_string, const char *modname)
 {
 	unsigned int flags = 0, mask = 0;
-	struct ddebug_query query;
+	struct ddebug_query query = {};
 #define MAXWORDS 9
Quoted 8 lines
Quoted 5 lines
lib/dynamic_debug.c +0 -1
 	int nwords, nfound;
 	char *words[MAXWORDS];
--

2.25.1

There's no need for this in stable kernels, please drop it from everywhere.

Dropped, thanks!

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

PATCH 53/60 bcache: fix super block seq numbers comparision in register_cache_set()

In reply to Sasha Levin

From: Coly Li <[email protected]>

[ Upstream commit 117f636ea695270fe492d0c0c9dfadc7a662af47 ]

In register_cache_set(), c is pointer to struct cache_set, and ca is pointer to struct cache, if ca->sb.seq > c->sb.seq, it means this registering cache has up to date version and other members, the in- memory version and other members should be updated to the newer value.

But current implementation makes a cache set only has a single cache device, so the above assumption works well except for a special case. The execption is when a cache device new created and both ca->sb.seq and c->sb.seq are 0, because the super block is never flushed out yet. In the location for the following if() check,
2156 if (ca->sb.seq > c->sb.seq) {

2157                 c->sb.version           = ca->sb.version;
2158                 memcpy(c->sb.set_uuid, ca->sb.set_uuid, 16);
2159                 c->sb.flags             = ca->sb.flags;
2160                 c->sb.seq               = ca->sb.seq;
2161                 pr_debug("set version = %llu\n", c->sb.version);

2162 }
c->sb.version is not initialized yet and valued 0. When ca->sb.seq is 0, the if() check will fail (because both values are 0), and the cache set version, set_uuid, flags and seq won't be updated.

The above problem is hiden for current code, because the bucket size is compatible among different super block version. And the next time when running cache set again, ca->sb.seq will be larger than 0 and cache set super block version will be updated properly.

But if the large bucket feature is enabled, sb->bucket_size is the low 16bits of the bucket size. For a power of 2 value, when the actual bucket size exceeds 16bit width, sb->bucket_size will always be 0. Then read_super_common() will fail because the if() check to is_power_of_2(sb->bucket_size) is false. This is how the long time hidden bug is triggered.

This patch modifies the if() check to the following way,
2156 if (ca->sb.seq > c->sb.seq || c->sb.seq == 0) {
Then cache set's version, set_uuid, flags and seq will always be updated corectly including for a new created cache device.

---
 drivers/md/bcache/super.c | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
drivers/md/bcache/super.c +8 -1
index a2e5a0fcd7d5c..7048370331c38 100644
@@ -2099,7 +2099,14 @@ static const char *register_cache_set(struct cache *ca)
 	    sysfs_create_link(&c->kobj, &ca->kobj, buf))
 		goto err;
 
-	if (ca->sb.seq > c->sb.seq) {
+	/*
+	 * A special case is both ca->sb.seq and c->sb.seq are 0,
+	 * such condition happens on a new created cache device whose
+	 * super block is never flushed yet. In this case c->sb.version
+	 * and other members should be updated too, otherwise we will
+	 * have a mistaken super block version in cache set.
+	 */
+	if (ca->sb.seq > c->sb.seq || c->sb.seq == 0) {
 		c->sb.version		= ca->sb.version;
 		memcpy(c->sb.set_uuid, ca->sb.set_uuid, 16);
 		c->sb.flags             = ca->sb.flags;
Signature · 1 line
2.25.1
author maintainer
#57 lore raw

PATCH 54/60 btrfs: allow btrfs_truncate_block() to fallback to nocow for data space reservation

In reply to Sasha Levin

From: Qu Wenruo <[email protected]>

[ Upstream commit 6d4572a9d71d5fc2affee0258d8582d39859188c ]

[BUG]
When the data space is exhausted, even if the inode has NOCOW attribute, we will still refuse to truncate unaligned range due to ENOSPC.

The following script can reproduce it pretty easily:
#!/bin/bash

dev=/dev/test/test
mnt=/mnt/btrfs

umount $dev &> /dev/null
umount $mnt &> /dev/null

mkfs.btrfs -f $dev -b 1G
mount -o nospace_cache $dev $mnt
touch $mnt/foobar
chattr +C $mnt/foobar

xfs_io -f -c "pwrite -b 4k 0 4k" $mnt/foobar > /dev/null xfs_io -f -c "pwrite -b 4k 0 1G" $mnt/padding &> /dev/null sync

xfs_io -c "fpunch 0 2k" $mnt/foobar
umount $mnt

Currently this will fail at the fpunch part.

[CAUSE]
Because btrfs_truncate_block() always reserves space without checking the NOCOW attribute.

Since the writeback path follows NOCOW bit, we only need to bother the space reservation code in btrfs_truncate_block().

[FIX]
Make btrfs_truncate_block() follow btrfs_buffered_write() to try to reserve data space first, and fall back to NOCOW check only when we don't have enough space.

Such always-try-reserve is an optimization introduced in btrfs_buffered_write(), to avoid expensive btrfs_check_can_nocow() call.

This patch will export check_can_nocow() as btrfs_check_can_nocow(), and use it in btrfs_truncate_block() to fix the problem.

---
 fs/btrfs/ctree.h |  2 ++
 fs/btrfs/file.c  | 12 ++++++------
 fs/btrfs/inode.c | 44 +++++++++++++++++++++++++++++++++++++-------
 3 files changed, 45 insertions(+), 13 deletions(-)
fs/btrfs/ctree.h +2 -0
index 09e6dff8a8f85..68bd89e3d4f09 100644
@@ -2982,6 +2982,8 @@ int btrfs_dirty_pages(struct inode *inode, struct page **pages,
 		      size_t num_pages, loff_t pos, size_t write_bytes,
 		      struct extent_state **cached);
 int btrfs_fdatawrite_range(struct inode *inode, loff_t start, loff_t end);
+int btrfs_check_can_nocow(struct btrfs_inode *inode, loff_t pos,
+			  size_t *write_bytes, bool nowait);
 
 /* tree-defrag.c */
 int btrfs_defrag_leaves(struct btrfs_trans_handle *trans,
fs/btrfs/file.c +6 -6
index 93244934d4f92..1e1af0ce70771 100644
@@ -1540,8 +1540,8 @@ lock_and_cleanup_extent_if_need(struct btrfs_inode *inode, struct page **pages,
 	return ret;
 }
 
-static noinline int check_can_nocow(struct btrfs_inode *inode, loff_t pos,
-				    size_t *write_bytes, bool nowait)
+int btrfs_check_can_nocow(struct btrfs_inode *inode, loff_t pos,
+			  size_t *write_bytes, bool nowait)
 {
 	struct btrfs_fs_info *fs_info = inode->root->fs_info;
 	struct btrfs_root *root = inode->root;
@@ -1656,8 +1656,8 @@ static noinline ssize_t btrfs_buffered_write(struct kiocb *iocb,
 		if (ret < 0) {
 			if ((BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
 						      BTRFS_INODE_PREALLOC)) &&
-			    check_can_nocow(BTRFS_I(inode), pos,
-					    &write_bytes, false) > 0) {
+			    btrfs_check_can_nocow(BTRFS_I(inode), pos,
+						  &write_bytes, false) > 0) {
 				/*
 				 * For nodata cow case, no need to reserve
 				 * data space.
@@ -1936,8 +1936,8 @@ static ssize_t btrfs_file_write_iter(struct kiocb *iocb,
 		 */
 		if (!(BTRFS_I(inode)->flags & (BTRFS_INODE_NODATACOW |
 					      BTRFS_INODE_PREALLOC)) ||
-		    check_can_nocow(BTRFS_I(inode), pos, &nocow_bytes,
-				    true) <= 0) {
+		    btrfs_check_can_nocow(BTRFS_I(inode), pos, &nocow_bytes,
+					  true) <= 0) {
 			inode_unlock(inode);
 			return -EAGAIN;
 		}
fs/btrfs/inode.c +37 -7

73 lines not on this page · raw message

Signature · 1 line
2.25.1
author maintainer
#58 lore raw

PATCH 55/60 btrfs: qgroup: free per-trans reserved space when a subvolume gets dropped

In reply to Sasha Levin

From: Qu Wenruo <[email protected]>

[ Upstream commit a3cf0e4342b6af9e6b34a4b913c630fbd03a82ea ]

[BUG]
Sometime fsstress could lead to qgroup warning for case like generic/013:

BTRFS warning (device dm-3): qgroup 0/259 has unreleased space, type 1 rsv 81920

  ------------[ cut here ]------------
  WARNING: CPU: 9 PID: 24535 at fs/btrfs/disk-io.c:4142 close_ctree+0x1dc/0x323 [btrfs]
  Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS 0.0.0 02/06/2015
  RIP: 0010:close_ctree+0x1dc/0x323 [btrfs]
  Call Trace:
   btrfs_put_super+0x15/0x17 [btrfs]
   generic_shutdown_super+0x72/0x110
   kill_anon_super+0x18/0x30
   btrfs_kill_super+0x17/0x30 [btrfs]
   deactivate_locked_super+0x3b/0xa0
   deactivate_super+0x40/0x50
   cleanup_mnt+0x135/0x190
   __cleanup_mnt+0x12/0x20
   task_work_run+0x64/0xb0
   __prepare_exit_to_usermode+0x1bc/0x1c0
   __syscall_return_slowpath+0x47/0x230
   do_syscall_64+0x64/0xb0
   entry_SYSCALL_64_after_hwframe+0x44/0xa9
  ---[ end trace 6c341cdf9b6cc3c1 ]---
  BTRFS error (device dm-3): qgroup reserved space leaked

While that subvolume 259 is no longer in that filesystem.

[CAUSE]
Normally per-trans qgroup reserved space is freed when a transaction is committed, in commit_fs_roots().

However for completely dropped subvolume, that subvolume is completely gone, thus is no longer in the fs_roots_radix, and its per-trans reserved qgroup will never be freed.

Since the subvolume is already gone, leaked per-trans space won't cause any trouble for end users.

[FIX]
Just call btrfs_qgroup_free_meta_all_pertrans() before a subvolume is completely dropped.

---
 fs/btrfs/extent-tree.c | 8 ++++++++
 1 file changed, 8 insertions(+)
fs/btrfs/extent-tree.c +8 -0
index 54a64d1e18c6b..7c86188b33d43 100644
@@ -5481,6 +5481,14 @@ int btrfs_drop_snapshot(struct btrfs_root *root, int update_ref, int for_reloc)
 		}
 	}
 
+	/*
+	 * This subvolume is going to be completely dropped, and won't be
+	 * recorded as dirty roots, thus pertrans meta rsv will not be freed at
+	 * commit transaction time.  So free it here manually.
+	 */
+	btrfs_qgroup_convert_reserved_meta(root, INT_MAX);
+	btrfs_qgroup_free_meta_all_pertrans(root);
+
 	if (test_bit(BTRFS_ROOT_IN_RADIX, &root->state))
 		btrfs_add_dropped_root(trans, root);
 	else
Signature · 1 line
2.25.1
author maintainer
#59 lore raw

PATCH 56/60 ACPICA: Do not increment operation_region reference counts for field units

In reply to Sasha Levin

From: Erik Kaneda <[email protected]>

[ Upstream commit 6a54ebae6d047c988a31f5ac5a64ab5cf83797a2 ]

ACPICA commit e17b28cfcc31918d0db9547b6b274b09c413eb70

Object reference counts are used as a part of ACPICA's garbage collection mechanism. This mechanism keeps track of references to heap-allocated structures such as the ACPI operand objects.

Recent server firmware has revealed that this reference count can overflow on large servers that declare many field units under the same operation_region. This occurs because each field unit declaration will add a reference count to the source operation_region.

This change solves the reference count overflow for operation_regions objects by preventing fieldunits from incrementing their operation_region's reference count. Each operation_region's reference count will not be changed by named objects declared under the Field operator. During namespace deletion, the operation_region namespace node will be deleted and each fieldunit will be deleted without touching the deleted operation_region object.

---
 drivers/acpi/acpica/exprep.c   | 4 ----
 drivers/acpi/acpica/utdelete.c | 6 +-----
 2 files changed, 1 insertion(+), 9 deletions(-)
drivers/acpi/acpica/exprep.c +0 -4
index a4e306690a21b..4a0f03157e082 100644
@@ -473,10 +473,6 @@ acpi_status acpi_ex_prep_field_value(struct acpi_create_field_info *info)
 				    (u8)access_byte_width;
 			}
 		}
-		/* An additional reference for the container */
-
-		acpi_ut_add_reference(obj_desc->field.region_obj);
-
 		ACPI_DEBUG_PRINT((ACPI_DB_BFIELD,
 				  "RegionField: BitOff %X, Off %X, Gran %X, Region %p\n",
 				  obj_desc->field.start_field_bit_offset,
drivers/acpi/acpica/utdelete.c +1 -5
index c365faf4e6cd4..4c0d4e4341961 100644
@@ -568,11 +568,6 @@ acpi_ut_update_object_reference(union acpi_operand_object *object, u16 action)
 			next_object = object->buffer_field.buffer_obj;
 			break;
 
-		case ACPI_TYPE_LOCAL_REGION_FIELD:
-
-			next_object = object->field.region_obj;
-			break;
-
 		case ACPI_TYPE_LOCAL_BANK_FIELD:
 
 			next_object = object->bank_field.bank_obj;
@@ -613,6 +608,7 @@ acpi_ut_update_object_reference(union acpi_operand_object *object, u16 action)
 			}
 			break;
 
+		case ACPI_TYPE_LOCAL_REGION_FIELD:
 		case ACPI_TYPE_REGION:
 		default:
 
Signature · 1 line
2.25.1
author maintainer
#60 lore raw

PATCH 57/60 io_uring: fix racy overflow count reporting

In reply to Sasha Levin

From: Pavel Begunkov <[email protected]>

[ Upstream commit b2bd1cf99f3e7c8fbf12ea07af2c6998e1209e25 ]

All ->cq_overflow modifications should be under completion_lock, otherwise it can report a wrong number to the userspace. Fix it in io_uring_cancel_files().

---
 fs/io_uring.c | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)
fs/io_uring.c +1 -2
index 80a612d1e12bb..d04fede20acb8 100644
@@ -7573,10 +7573,9 @@ static void io_uring_cancel_files(struct io_ring_ctx *ctx,
 				clear_bit(0, &ctx->sq_check_overflow);
 				clear_bit(0, &ctx->cq_check_overflow);
 			}
-			spin_unlock_irq(&ctx->completion_lock);
-
 			WRITE_ONCE(ctx->rings->cq_overflow,
 				atomic_inc_return(&ctx->cached_cq_overflow));
+			spin_unlock_irq(&ctx->completion_lock);
 
 			/*
 			 * Put inflight ref and overflow ref. If that's
Signature · 1 line
2.25.1
author maintainer
#61 lore raw

PATCH 58/60 io_uring: fix stalled deferred requests

In reply to Sasha Levin

From: Pavel Begunkov <[email protected]>

[ Upstream commit dd9dfcdf5a603680458f5e7b0d2273c66e5417db ]

Always do io_commit_cqring() after completing a request, even if it was accounted as overflowed on the CQ side. Failing to do that may lead to not to pushing deferred requests when needed, and so stalling the whole ring.

---
 fs/io_uring.c | 1 +
 1 file changed, 1 insertion(+)
fs/io_uring.c +1 -0
index d04fede20acb8..5d33c05de02e5 100644
@@ -7575,6 +7575,7 @@ static void io_uring_cancel_files(struct io_ring_ctx *ctx,
 			}
 			WRITE_ONCE(ctx->rings->cq_overflow,
 				atomic_inc_return(&ctx->cached_cq_overflow));
+			io_commit_cqring(ctx);
 			spin_unlock_irq(&ctx->completion_lock);
 
 			/*
Signature · 1 line
2.25.1
author maintainer
#62 lore raw

PATCH 59/60 crypto: caam - silence .setkey in case of bad key length

In reply to Sasha Levin

From: Horia Geantă <[email protected]>

[ Upstream commit da6a66853a381864f4b040832cf11f0dbba0a097 ]

In case of bad key length, driver emits "key size mismatch" messages, but only for xts(aes) algorithms.

Reduce verbosity by making them visible only when debugging. This way crypto fuzz testing log cleans up a bit.

---
 drivers/crypto/caam/caamalg.c     | 2 +-
 drivers/crypto/caam/caamalg_qi.c  | 2 +-
 drivers/crypto/caam/caamalg_qi2.c | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)
drivers/crypto/caam/caamalg.c +1 -1
index b2f9882bc010f..bf90a4fcabd1f 100644
@@ -838,7 +838,7 @@ static int xts_skcipher_setkey(struct crypto_skcipher *skcipher, const u8 *key,
 	u32 *desc;
 
 	if (keylen != 2 * AES_MIN_KEY_SIZE  && keylen != 2 * AES_MAX_KEY_SIZE) {
-		dev_err(jrdev, "key size mismatch\n");
+		dev_dbg(jrdev, "key size mismatch\n");
 		return -EINVAL;
 	}
 
drivers/crypto/caam/caamalg_qi.c +1 -1
index 27e36bdf6163b..315d53499ce85 100644
@@ -728,7 +728,7 @@ static int xts_skcipher_setkey(struct crypto_skcipher *skcipher, const u8 *key,
 	int ret = 0;
 
 	if (keylen != 2 * AES_MIN_KEY_SIZE  && keylen != 2 * AES_MAX_KEY_SIZE) {
-		dev_err(jrdev, "key size mismatch\n");
+		dev_dbg(jrdev, "key size mismatch\n");
 		return -EINVAL;
 	}
 
drivers/crypto/caam/caamalg_qi2.c +1 -1
index 28669cbecf77c..e1b6bc6ef091b 100644
@@ -1058,7 +1058,7 @@ static int xts_skcipher_setkey(struct crypto_skcipher *skcipher, const u8 *key,
 	u32 *desc;
 
 	if (keylen != 2 * AES_MIN_KEY_SIZE  && keylen != 2 * AES_MAX_KEY_SIZE) {
-		dev_err(dev, "key size mismatch\n");
+		dev_dbg(dev, "key size mismatch\n");
 		return -EINVAL;
 	}
 
Signature · 1 line
2.25.1
author maintainer
#63 lore raw

PATCH 60/60 drm/msm: ratelimit crtc event overflow error

In reply to Sasha Levin

From: Rob Clark <[email protected]>

[ Upstream commit 5e16372b5940b1fecc3cc887fc02a50ba148d373 ]

This can happen a lot when things go pear shaped. Lets not flood dmesg when this happens.

---
 drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/gpu/drm/msm/disp/dpu1/dpu_crtc.c +1 -1
index 17448505a9b5f..d263d6e69bf12 100644
@@ -386,7 +386,7 @@ static void dpu_crtc_frame_event_cb(void *data, u32 event)
 	spin_unlock_irqrestore(&dpu_crtc->spin_lock, flags);
 
 	if (!fevent) {
-		DRM_ERROR("crtc%d event %d overflow\n", crtc->base.id, event);
+		DRM_ERROR_RATELIMITED("crtc%d event %d overflow\n", crtc->base.id, event);
 		return;
 	}
 
Signature · 6 lines
2.25.1

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