PCI: Init driver override spinlock in new_id_store()
PATCH 1/1 PCI: Init driver override spinlock in new_id_store()
When setting new_id of a PCI device driver using sysfs a lockdep splat occurs. This is because the new_id function new_id_store() checks for driver_override by creating temporary pci_dev structs.
Since the newly added driver_override spinlock is not init for temporary pci_dev structs, the lockdep splat complains about it.
Init the driver_override spinlock to fix this.
[ 4.464296] INFO: trying to register non-static key. [ 4.466207] The code is fine but needs lockdep annotation, or maybe [ 4.468487] you didn't initialize this object before use? [ 4.470486] turning off the locking correctness validator. [ 4.472494] CPU: 2 UID: 0 PID: 177 Comm: liveupdate-iomm Not tainted 7.0.0+ #9 PREEMPT(full) [ 4.472497] Hardware name: QEMU Standard PC (Q35 + ICH9, 2009), BIOS rel-1.16.3-0-ga6ed6b701f0a-prebuilt.qemu.org 04/01/2014 [ 4.472498] Call Trace: [ 4.472501] <TASK> [ 4.472503] dump_stack_lvl+0x5d/0x80 [ 4.472508] register_lock_class+0x77e/0x790 [ 4.472512] ? check_prev_add+0xf4/0xd30 [ 4.472513] ? check_prev_add+0xf4/0xd30 [ 4.472515] __lock_acquire+0x3b4/0x1b80 [ 4.472518] ? raw_irqentry_exit_cond_resched+0x20/0x50 [ 4.472556] lock_acquire+0xbf/0x2e0 [ 4.472558] ? pci_match_device+0x24/0x180 [ 4.472563] _raw_spin_lock+0x30/0x40 [ 4.472567] ? pci_match_device+0x24/0x180 [ 4.472569] pci_match_device+0x24/0x180 [ 4.472571] new_id_store+0x189/0x1d0 [ 4.472574] kernfs_fop_write_iter+0x14f/0x210 [ 4.472578] vfs_write+0x263/0x5e0 [ 4.472583] ksys_write+0x79/0xf0 [ 4.472585] do_syscall_64+0x117/0xf80 [ 4.472587] ? clear_bhb_loop+0x40/0x90 [ 4.472590] entry_SYSCALL_64_after_hwframe+0x77/0x7f [ 4.472592] RIP: 0033:0x7fe7bc0beeb2 [ 4.472595] Code: 18 41 8b 93 08 03 00 00 59 5e 48 83 f8 fc 75 1a 83 e2 39 83 fa 08 75 12 e8 2b ff ff ff 0f 1f 00 49 89 ca 48 8b 44 24 20 0f 05 <48> 83 c4 18 c3 66 0f 1f 84 00 00 00 00 00 48 83 ec 10 ff 74 24 18 [ 4.472596] RSP: 002b:00007ffcdbb1ed80 EFLAGS: 00000202 ORIG_RAX: 0000000000000001 [ 4.472598] RAX: ffffffffffffffda RBX: 00007fe7bc2415c0 RCX: 00007fe7bc0beeb2 [ 4.472599] RDX: 000000000000000a RSI: 00005580955433c0 RDI: 0000000000000001 [ 4.472600] RBP: 000000000000000a R08: 0000000000000000 R09: 0000000000000000 [ 4.472601] R10: 0000000000000000 R11: 0000000000000202 R12: 000000000000000a [ 4.472602] R13: 00005580955433c0 R14: 000055805932493e R15: 0000558095545490 [ 4.472605] </TASK>
--- drivers/pci/pci-driver.c | 5 +++++ 1 file changed, 5 insertions(+)
drivers/pci/pci-driver.c +5 -0
@@ -215,6 +215,11 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_device = subdevice; pdev->class = class; +/* +* Initialize the embedded struct device driver_override lock to +* avoid the lockdep errors. +*/ +spin_lock_init (&pdev->dev.driver_override.lock);if (pci_match_device (pdrv, pdev)) retval = -EEXIST ;
base-commit: 254f49634ee16a731174d2ae34bc50bd5f45e731
Signature · 1 line
2.54.0.545.g6539524ca2-goog
Quoted 1 line · Samiullah Khawaja
Fixes cb3d1049f4ea ("driver core: generalize driver_override in struct device")
I don't think anything is wrong with this commit, and it seems unrelated.
Quoted 1 line
Fixes 10a4206a2401 ("PCI: use generic driver_override infrastructure")
I'm also not sure that this one contains the root cause, despite revealing the actual issue, more below.
Quoted 10 lines
Signed-off-by Samiullah Khawaja <[email protected]>--- drivers/pci/pci-driver.c | 5 +++++ 1 file changed, 5 insertions(+)drivers/pci/pci-driver.c +0
-0
drivers/pci/pci-driver.c static ssize_t new_id_store(struct device_driver *driver, const char *buf, · line 222
Quoted 8 lines
drivers/pci/pci-driver.c +5
-0@@ -215,6 +215,11 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_device = subdevice; pdev->class = class; +/* +* Initialize the embedded struct device driver_override lock to +* avoid the lockdep errors. +*/ +spin_lock_init (&pdev->dev.driver_override.lock);
Can't we just call device_initialize() and set pdev->dev.release to a new function that just calls kfree()?
This way nothing of that kind can ever happen again; it is hard to predict that a device is used without ever being initialized.
Quoted 7 lines
drivers/pci/pci-driver.c +0
-0if (pci_match_device (pdrv, pdev)) retval = -EEXIST ;base-commit:
254f49634ee16a731174d2ae34bc50bd5f45e731Signature · 1 line
2.54.0.545.g6539524ca2-goog
Quoted 4 lines · Danilo Krummrich
Quoted 1 line · Samiullah Khawaja
Fixes cb3d1049f4ea ("driver core: generalize driver_override in struct device")I don't think anything is wrong with this commit, and it seems unrelated.
I will remove this one.
Quoted 5 lines
Quoted 2 lines
Fixes 10a4206a2401 ("PCI: use generic driver_override infrastructure")I'm also not sure that this one contains the root cause, despite revealing the actual issue, more below.
I see your point. I understand the actual issue is that the temporary pci_dev struct was not init and that was added in 2014. I added this since this is where the regression happened.
Quoted 23 lines
Quoted 11 lines
Signed-off-by Samiullah Khawaja <[email protected]>--- drivers/pci/pci-driver.c | 5 +++++ 1 file changed, 5 insertions(+)drivers/pci/pci-driver.c +0
-0drivers/pci/pci-driver.c static ssize_t new_id_store(struct device_driver *driver, const char *buf, · line 222
Quoted 8 lines
drivers/pci/pci-driver.c +5
-0@@ -215,6 +215,11 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_device = subdevice; pdev->class = class; +/* +* Initialize the embedded struct device driver_override lock to +* avoid the lockdep errors. +*/ +spin_lock_init (&pdev->dev.driver_override.lock);Can't we just call
device_initialize()and setpdev->dev.releaseto a new function that just callskfree()?
Quoted 2 lines
This way nothing of that kind can ever happen again; it is hard to predict that a device is used without ever being initialized.
Agreed.
yes we can set the release function and call device_initialize().
Should I send v2 with something like following:
drivers/pci/pci-driver.c +9 -1
@@ -179,6 +179,11 @@ static const struct pci_device_id *pci_match_device(struct pci_driver *drv,return NULL ; } +static void _release_temp_device (struct device *dev) +{ +kfree (to_pci_dev (dev)); +} +/** * new_id_store - sysfs frontend to pci_add_dynid() * @driver: target device driver
@@ -214,11 +219,14 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_vendor = subvendor; pdev->subsystem_device = subdevice; pdev->class = class; + pdev->dev.release = _release_temp_device; +/* Initialize the embedded struct device. */ +device_initialize (&pdev->dev);if (pci_match_device (pdrv, pdev)) retval = -EEXIST ; -kfree (pdev); +put_device (&pdev->dev);if (retval)return retval;
Quoted 8 lines
Quoted 8 lines
drivers/pci/pci-driver.c +0
-0if (pci_match_device (pdrv, pdev)) retval = -EEXIST ;base-commit:
254f49634ee16a731174d2ae34bc50bd5f45e731--
2.54.0.545.g6539524ca2-goog
Quoted 8 lines · Samiullah Khawaja
Quoted 4 lines · Danilo Krummrich
Quoted 1 line · Samiullah Khawaja
Fixes cb3d1049f4ea ("driver core: generalize driver_override in struct device")I don't think anything is wrong with this commit, and it seems unrelated.
I will remove this one.
Quoted 1 line
Quoted 8 lines
Quoted 4 lines
Quoted 1 line
Fixes 10a4206a2401 ("PCI: use generic driver_override infrastructure")I'm also not sure that this one contains the root cause, despite revealing the actual issue, more below.
I see your point. I understand the actual issue is that the temporary
pci_devstruct was not init and that was added in 2014. I added this since this is where the regression happened.
Yeah, I did not mean to say that this tag is wrong -- it is where the regression happened. But it may be worth adding both, this one and the one that did omit device_initialize() in the first place.
More in general, this can matter because it influences how (far) it is backported in stable trees.
While rather unlikely in this case, there could be a subsequent fix that is incompatible with device_initialize() not being called that may be backported event further.
This would not regress in recent trees as it would have been fixed there already, but might regress in an ancient stable tree, that did not get this fix.
Quoted 50 lines
Quoted 22 lines
Quoted 10 lines
Signed-off-by Samiullah Khawaja <[email protected]>--- drivers/pci/pci-driver.c | 5 +++++ 1 file changed, 5 insertions(+)drivers/pci/pci-driver.c +0
-0drivers/pci/pci-driver.c static ssize_t new_id_store(struct device_driver *driver, const char *buf, · line 222
Quoted 8 lines
drivers/pci/pci-driver.c +5
-0@@ -215,6 +215,11 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_device = subdevice; pdev->class = class; +/* +* Initialize the embedded struct device driver_override lock to +* avoid the lockdep errors. +*/ +spin_lock_init (&pdev->dev.driver_override.lock);Can't we just call
device_initialize()and setpdev->dev.releaseto a new function that just callskfree()?Quoted 2 lines
This way nothing of that kind can ever happen again; it is hard to predict that a device is used without ever being initialized.
Agreed.
yes we can set the release function and call
device_initialize().Should I send v2 with something like following:
drivers/pci/pci-driver.c +5
-0@@ -179,6 +179,11 @@ static const struct pci_device_id *pci_match_device(struct pci_driver *drv,return NULL ; } +static void _release_temp_device (struct device *dev) +{ +kfree (to_pci_dev (dev)); +} +/** * new_id_store - sysfs frontend to pci_add_dynid() * @driver: target device driver @@ -214,11 +219,14 @@ static ssize_t new_id_store(struct device_driver *driver, const char *buf, pdev->subsystem_vendor = subvendor; pdev->subsystem_device = subdevice;
drivers/pci/pci-driver.c static ssize_t new_id_store(struct device_driver *driver, const char *buf, · line 229
Quoted 10 lines
drivers/pci/pci-driver.c +4
-1pdev->class = class; + pdev->dev.release = _release_temp_device; +/* Initialize the embedded struct device. */ +device_initialize (&pdev->dev);if (pci_match_device (pdrv, pdev)) retval = -EEXIST ; -kfree (pdev); +put_device (&pdev->dev);
Overall Looks good. Not sure about the _release_temp_device() name and I think the comment above device_initialize() may be unnecessary, but I'm pretty sure the PCI maintainers will have an opinion on that. :)