[PATCH 0/1] coresight: Fix for v5.3-rc3

[PATCH 0/1] coresight: Fix for v5.3-rc3

[Mathieu Poirier]

Good morning Greg,

Here is a fix I'd like you to consider for this cycle.  Do you think you
could apply it to driver-core-next rather than char-misc-next?  My current
coresight branch is based on driver-core-next to pick up Suzuki's
generic device lookup helpers patchset [1]. Applying it to char-misc-next
will create two different signature for the same patch, something that
gives Stephen a hard time when building the linux-next tree.

I also think this patch should go in stable but haven't marked it as such since
it doesn't apply to any of the stable trees.  Once it is part of v5.3 I intend 
to send a new version of the patch that does apply cleanly to those trees.  Let
me know if you want me to proceed differently. 

Thanks,
Mathieu

[1]. https://www.spinics.net/lists/dri-devel/msg219674.html


Suzuki K Poulose (1):
  coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute

 drivers/hwtracing/coresight/coresight-etm-perf.c | 1 +
 1 file changed, 1 insertion(+)

-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

[PATCH 1/1] coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute

[Mathieu Poirier]

From: Suzuki K Poulose <suzuki.poulose@arm.com>

While running the linux-next with CONFIG_DEBUG_LOCKS_ALLOC enabled,
I get the following splat.

 BUG: key ffffcb5636929298 has not been registered!
 ------------[ cut here ]------------
 DEBUG_LOCKS_WARN_ON(1)
 WARNING: CPU: 1 PID: 53 at kernel/locking/lockdep.c:3669 lockdep_init_map+0x164/0x1f0
 CPU: 1 PID: 53 Comm: kworker/1:1 Tainted: G        W         5.2.0-next-20190712-00015-g00ad4634222e-dirty #603
 Workqueue: events amba_deferred_retry_func
 pstate: 60c00005 (nZCv daif +PAN +UAO)
 pc : lockdep_init_map+0x164/0x1f0
 lr : lockdep_init_map+0x164/0x1f0

 [ trimmed ]

 Call trace:
  lockdep_init_map+0x164/0x1f0
  __kernfs_create_file+0x9c/0x158
  sysfs_add_file_mode_ns+0xa8/0x1d0
  sysfs_add_file_to_group+0x88/0xd8
  etm_perf_add_symlink_sink+0xcc/0x138
  coresight_register+0x110/0x280
  tmc_probe+0x160/0x420

 [ trimmed ]

 ---[ end trace ab4cc669615ba1b0 ]---

Fix this by initialising the dynamically allocated attribute properly.

Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Fixes: bb8e370bdc141ddf ("coresight: perf: Add "sinks" group to PMU directory")
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
[Fixed a typograhic error in the changelog]
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
---
 drivers/hwtracing/coresight/coresight-etm-perf.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/hwtracing/coresight/coresight-etm-perf.c b/drivers/hwtracing/coresight/coresight-etm-perf.c
index 5c1ca0df5cb0..84f1dcb69827 100644
--- a/drivers/hwtracing/coresight/coresight-etm-perf.c
+++ b/drivers/hwtracing/coresight/coresight-etm-perf.c
@@ -544,6 +544,7 @@ int etm_perf_add_symlink_sink(struct coresight_device *csdev)
 	/* See function coresight_get_sink_by_id() to know where this is used */
 	hash = hashlen_hash(hashlen_string(NULL, name));
 
+	sysfs_attr_init(&ea->attr.attr);
 	ea->attr.attr.name = devm_kstrdup(dev, name, GFP_KERNEL);
 	if (!ea->attr.attr.name)
 		return -ENOMEM;
-- 
2.17.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 0/1] coresight: Fix for v5.3-rc3

[Greg KH]

On Thu, Aug 01, 2019 at 11:23:22AM -0600, Mathieu Poirier wrote:
> Good morning Greg,
> 
> Here is a fix I'd like you to consider for this cycle.  Do you think you
> could apply it to driver-core-next rather than char-misc-next?

All of my -next branches are for 5.4-rc1, not 5.3 (i.e. the "next"
kernel).

So either one of them isn't going to matter to you for 5.3-final.

> My current
> coresight branch is based on driver-core-next to pick up Suzuki's
> generic device lookup helpers patchset [1]. Applying it to char-misc-next
> will create two different signature for the same patch, something that
> gives Stephen a hard time when building the linux-next tree.

Why would Suzuki's patch matter for 5.3-final?

> I also think this patch should go in stable but haven't marked it as such since
> it doesn't apply to any of the stable trees.  Once it is part of v5.3 I intend 
> to send a new version of the patch that does apply cleanly to those trees.  Let
> me know if you want me to proceed differently. 
> 
> Thanks,
> Mathieu
> 
> [1]. https://www.spinics.net/lists/dri-devel/msg219674.html
> 
> 
> Suzuki K Poulose (1):
>   coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute
> 
>  drivers/hwtracing/coresight/coresight-etm-perf.c | 1 +
>  1 file changed, 1 insertion(+)

Why would this patch depend on anything in any of my trees?

totally confused,

greg k-h

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 1/1] coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute

[Greg KH]

On Thu, Aug 01, 2019 at 11:23:23AM -0600, Mathieu Poirier wrote:
> From: Suzuki K Poulose <suzuki.poulose@arm.com>
> 
> While running the linux-next with CONFIG_DEBUG_LOCKS_ALLOC enabled,
> I get the following splat.
> 
>  BUG: key ffffcb5636929298 has not been registered!
>  ------------[ cut here ]------------
>  DEBUG_LOCKS_WARN_ON(1)
>  WARNING: CPU: 1 PID: 53 at kernel/locking/lockdep.c:3669 lockdep_init_map+0x164/0x1f0
>  CPU: 1 PID: 53 Comm: kworker/1:1 Tainted: G        W         5.2.0-next-20190712-00015-g00ad4634222e-dirty #603
>  Workqueue: events amba_deferred_retry_func
>  pstate: 60c00005 (nZCv daif +PAN +UAO)
>  pc : lockdep_init_map+0x164/0x1f0
>  lr : lockdep_init_map+0x164/0x1f0
> 
>  [ trimmed ]
> 
>  Call trace:
>   lockdep_init_map+0x164/0x1f0
>   __kernfs_create_file+0x9c/0x158
>   sysfs_add_file_mode_ns+0xa8/0x1d0
>   sysfs_add_file_to_group+0x88/0xd8
>   etm_perf_add_symlink_sink+0xcc/0x138
>   coresight_register+0x110/0x280
>   tmc_probe+0x160/0x420
> 
>  [ trimmed ]
> 
>  ---[ end trace ab4cc669615ba1b0 ]---
> 
> Fix this by initialising the dynamically allocated attribute properly.
> 
> Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
> Fixes: bb8e370bdc141ddf ("coresight: perf: Add "sinks" group to PMU directory")

That should be:
Fixes: bb8e370bdc14 ("coresight: perf: Add "sinks" group to PMU directory")

I can go fix it up...


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 0/1] coresight: Fix for v5.3-rc3

[Mathieu Poirier]

On Thu, 1 Aug 2019 at 12:17, Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Thu, Aug 01, 2019 at 11:23:22AM -0600, Mathieu Poirier wrote:
> > Good morning Greg,
> >
> > Here is a fix I'd like you to consider for this cycle.  Do you think you
> > could apply it to driver-core-next rather than char-misc-next?
>
> All of my -next branches are for 5.4-rc1, not 5.3 (i.e. the "next"
> kernel).
>
> So either one of them isn't going to matter to you for 5.3-final.
>
> > My current
> > coresight branch is based on driver-core-next to pick up Suzuki's
> > generic device lookup helpers patchset [1]. Applying it to char-misc-next
> > will create two different signature for the same patch, something that
> > gives Stephen a hard time when building the linux-next tree.
>
> Why would Suzuki's patch matter for 5.3-final?

There was a similar situation during the 5.2 cycle [1].  Here too we
can fix a problem that is present in 5.3 rather than wait for 5.4.

[1]. https://www.spinics.net/lists/arm-kernel/msg736274.html

>
> > I also think this patch should go in stable but haven't marked it as such since
> > it doesn't apply to any of the stable trees.  Once it is part of v5.3 I intend
> > to send a new version of the patch that does apply cleanly to those trees.  Let
> > me know if you want me to proceed differently.
> >
> > Thanks,
> > Mathieu
> >
> > [1]. https://www.spinics.net/lists/dri-devel/msg219674.html
> >
> >
> > Suzuki K Poulose (1):
> >   coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute
> >
> >  drivers/hwtracing/coresight/coresight-etm-perf.c | 1 +
> >  1 file changed, 1 insertion(+)
>
> Why would this patch depend on anything in any of my trees?

I send you patches for inclusion in the next cycle and as such I
thought it should be the same for fixes in the current cycle.  If that
is not the case, should I send them directly to Linus?

Thanks,
Mathieu

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 1/1] coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute

[Mathieu Poirier]

On Thu, 1 Aug 2019 at 12:50, Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Thu, Aug 01, 2019 at 11:23:23AM -0600, Mathieu Poirier wrote:
> > From: Suzuki K Poulose <suzuki.poulose@arm.com>
> >
> > While running the linux-next with CONFIG_DEBUG_LOCKS_ALLOC enabled,
> > I get the following splat.
> >
> >  BUG: key ffffcb5636929298 has not been registered!
> >  ------------[ cut here ]------------
> >  DEBUG_LOCKS_WARN_ON(1)
> >  WARNING: CPU: 1 PID: 53 at kernel/locking/lockdep.c:3669 lockdep_init_map+0x164/0x1f0
> >  CPU: 1 PID: 53 Comm: kworker/1:1 Tainted: G        W         5.2.0-next-20190712-00015-g00ad4634222e-dirty #603
> >  Workqueue: events amba_deferred_retry_func
> >  pstate: 60c00005 (nZCv daif +PAN +UAO)
> >  pc : lockdep_init_map+0x164/0x1f0
> >  lr : lockdep_init_map+0x164/0x1f0
> >
> >  [ trimmed ]
> >
> >  Call trace:
> >   lockdep_init_map+0x164/0x1f0
> >   __kernfs_create_file+0x9c/0x158
> >   sysfs_add_file_mode_ns+0xa8/0x1d0
> >   sysfs_add_file_to_group+0x88/0xd8
> >   etm_perf_add_symlink_sink+0xcc/0x138
> >   coresight_register+0x110/0x280
> >   tmc_probe+0x160/0x420
> >
> >  [ trimmed ]
> >
> >  ---[ end trace ab4cc669615ba1b0 ]---
> >
> > Fix this by initialising the dynamically allocated attribute properly.
> >
> > Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
> > Fixes: bb8e370bdc141ddf ("coresight: perf: Add "sinks" group to PMU directory")
>
> That should be:
> Fixes: bb8e370bdc14 ("coresight: perf: Add "sinks" group to PMU directory")

Apologies - I will catch those from hereon.

>
> I can go fix it up...

Or I can, depending on if you're going to pick the patch as a fix for
5.3 or you'll wait for the 5.4 cycle.

>

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 1/1] coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute

[Greg KH]

On Thu, Aug 01, 2019 at 02:20:39PM -0600, Mathieu Poirier wrote:
> On Thu, 1 Aug 2019 at 12:50, Greg KH <gregkh@linuxfoundation.org> wrote:
> >
> > On Thu, Aug 01, 2019 at 11:23:23AM -0600, Mathieu Poirier wrote:
> > > From: Suzuki K Poulose <suzuki.poulose@arm.com>
> > >
> > > While running the linux-next with CONFIG_DEBUG_LOCKS_ALLOC enabled,
> > > I get the following splat.
> > >
> > >  BUG: key ffffcb5636929298 has not been registered!
> > >  ------------[ cut here ]------------
> > >  DEBUG_LOCKS_WARN_ON(1)
> > >  WARNING: CPU: 1 PID: 53 at kernel/locking/lockdep.c:3669 lockdep_init_map+0x164/0x1f0
> > >  CPU: 1 PID: 53 Comm: kworker/1:1 Tainted: G        W         5.2.0-next-20190712-00015-g00ad4634222e-dirty #603
> > >  Workqueue: events amba_deferred_retry_func
> > >  pstate: 60c00005 (nZCv daif +PAN +UAO)
> > >  pc : lockdep_init_map+0x164/0x1f0
> > >  lr : lockdep_init_map+0x164/0x1f0
> > >
> > >  [ trimmed ]
> > >
> > >  Call trace:
> > >   lockdep_init_map+0x164/0x1f0
> > >   __kernfs_create_file+0x9c/0x158
> > >   sysfs_add_file_mode_ns+0xa8/0x1d0
> > >   sysfs_add_file_to_group+0x88/0xd8
> > >   etm_perf_add_symlink_sink+0xcc/0x138
> > >   coresight_register+0x110/0x280
> > >   tmc_probe+0x160/0x420
> > >
> > >  [ trimmed ]
> > >
> > >  ---[ end trace ab4cc669615ba1b0 ]---
> > >
> > > Fix this by initialising the dynamically allocated attribute properly.
> > >
> > > Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
> > > Fixes: bb8e370bdc141ddf ("coresight: perf: Add "sinks" group to PMU directory")
> >
> > That should be:
> > Fixes: bb8e370bdc14 ("coresight: perf: Add "sinks" group to PMU directory")
> 
> Apologies - I will catch those from hereon.
> 
> >
> > I can go fix it up...
> 
> Or I can, depending on if you're going to pick the patch as a fix for
> 5.3 or you'll wait for the 5.4 cycle.

It's already queued up for my next round of patches to Linus to be
during the 5.3-rc cycle.

thanks,

greg k-h

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 0/1] coresight: Fix for v5.3-rc3

[Greg KH]

On Thu, Aug 01, 2019 at 02:16:46PM -0600, Mathieu Poirier wrote:
> On Thu, 1 Aug 2019 at 12:17, Greg KH <gregkh@linuxfoundation.org> wrote:
> >
> > On Thu, Aug 01, 2019 at 11:23:22AM -0600, Mathieu Poirier wrote:
> > > Good morning Greg,
> > >
> > > Here is a fix I'd like you to consider for this cycle.  Do you think you
> > > could apply it to driver-core-next rather than char-misc-next?
> >
> > All of my -next branches are for 5.4-rc1, not 5.3 (i.e. the "next"
> > kernel).
> >
> > So either one of them isn't going to matter to you for 5.3-final.
> >
> > > My current
> > > coresight branch is based on driver-core-next to pick up Suzuki's
> > > generic device lookup helpers patchset [1]. Applying it to char-misc-next
> > > will create two different signature for the same patch, something that
> > > gives Stephen a hard time when building the linux-next tree.
> >
> > Why would Suzuki's patch matter for 5.3-final?
> 
> There was a similar situation during the 5.2 cycle [1].  Here too we
> can fix a problem that is present in 5.3 rather than wait for 5.4.
> 
> [1]. https://www.spinics.net/lists/arm-kernel/msg736274.html

But that has nothing to do with Suzuki's patch that is in my driver core
tree.  I still fail to see the dependency here.

> > > I also think this patch should go in stable but haven't marked it as such since
> > > it doesn't apply to any of the stable trees.  Once it is part of v5.3 I intend
> > > to send a new version of the patch that does apply cleanly to those trees.  Let
> > > me know if you want me to proceed differently.
> > >
> > > Thanks,
> > > Mathieu
> > >
> > > [1]. https://www.spinics.net/lists/dri-devel/msg219674.html
> > >
> > >
> > > Suzuki K Poulose (1):
> > >   coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute
> > >
> > >  drivers/hwtracing/coresight/coresight-etm-perf.c | 1 +
> > >  1 file changed, 1 insertion(+)
> >
> > Why would this patch depend on anything in any of my trees?
> 
> I send you patches for inclusion in the next cycle and as such I
> thought it should be the same for fixes in the current cycle.  If that
> is not the case, should I send them directly to Linus?

You can send me fixes to forward on to Linus for this current cycle, if
they are not depending on patches that are only for the -next release.

I always keep 2 branches in my git trees:
	-linus : patches for Linus this release cycle
	-next  : patches for Linus the next release cycle

If you have bugfixes, make them against my -linus branch and send them
on.  Odds are they don't have dependancies other than what is in Linus's
tree, so that's fine.

If you have patches for the next cycle, send them against my -next
branch.

thanks,

greg k-h

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

Re: [PATCH 0/1] coresight: Fix for v5.3-rc3

[Mathieu Poirier]

On Fri, 2 Aug 2019 at 03:10, Greg KH <gregkh@linuxfoundation.org> wrote:
>
> On Thu, Aug 01, 2019 at 02:16:46PM -0600, Mathieu Poirier wrote:
> > On Thu, 1 Aug 2019 at 12:17, Greg KH <gregkh@linuxfoundation.org> wrote:
> > >
> > > On Thu, Aug 01, 2019 at 11:23:22AM -0600, Mathieu Poirier wrote:
> > > > Good morning Greg,
> > > >
> > > > Here is a fix I'd like you to consider for this cycle.  Do you think you
> > > > could apply it to driver-core-next rather than char-misc-next?
> > >
> > > All of my -next branches are for 5.4-rc1, not 5.3 (i.e. the "next"
> > > kernel).
> > >
> > > So either one of them isn't going to matter to you for 5.3-final.
> > >
> > > > My current
> > > > coresight branch is based on driver-core-next to pick up Suzuki's
> > > > generic device lookup helpers patchset [1]. Applying it to char-misc-next
> > > > will create two different signature for the same patch, something that
> > > > gives Stephen a hard time when building the linux-next tree.
> > >
> > > Why would Suzuki's patch matter for 5.3-final?
> >
> > There was a similar situation during the 5.2 cycle [1].  Here too we
> > can fix a problem that is present in 5.3 rather than wait for 5.4.
> >
> > [1]. https://www.spinics.net/lists/arm-kernel/msg736274.html
>
> But that has nothing to do with Suzuki's patch that is in my driver core
> tree.  I still fail to see the dependency here.

There is indeed no correlation between the fix and the patches in the
core tree.  When writing the original email I didn't know in what tree
you would end up applying the patch and feared we'd end up having the
same patch in two different tree.  Thanks to your explanation below
this won't happen.

>
> > > > I also think this patch should go in stable but haven't marked it as such since
> > > > it doesn't apply to any of the stable trees.  Once it is part of v5.3 I intend
> > > > to send a new version of the patch that does apply cleanly to those trees.  Let
> > > > me know if you want me to proceed differently.
> > > >
> > > > Thanks,
> > > > Mathieu
> > > >
> > > > [1]. https://www.spinics.net/lists/dri-devel/msg219674.html
> > > >
> > > >
> > > > Suzuki K Poulose (1):
> > > >   coresight: Fix DEBUG_LOCKS_WARN_ON for uninitialized attribute
> > > >
> > > >  drivers/hwtracing/coresight/coresight-etm-perf.c | 1 +
> > > >  1 file changed, 1 insertion(+)
> > >
> > > Why would this patch depend on anything in any of my trees?
> >
> > I send you patches for inclusion in the next cycle and as such I
> > thought it should be the same for fixes in the current cycle.  If that
> > is not the case, should I send them directly to Linus?
>
> You can send me fixes to forward on to Linus for this current cycle, if
> they are not depending on patches that are only for the -next release.
>
> I always keep 2 branches in my git trees:
>         -linus : patches for Linus this release cycle
>         -next  : patches for Linus the next release cycle
>
> If you have bugfixes, make them against my -linus branch and send them
> on.  Odds are they don't have dependancies other than what is in Linus's
> tree, so that's fine.

Perfect, that's the link I was missing.

Thanks,
Mathieu

>
> If you have patches for the next cycle, send them against my -next
> branch.
>
> thanks,
>
> greg k-h

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel