linux-next: manual merge of the driver-core tree with the driver-core.current tree

linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1022 bytes --]

Hi Greg,

Today's linux-next merge of the driver-core tree got a conflict in
drivers/base/cpu.c between commit 2885e25c422f ("driver core: cpu: remove
kernel warning when removing a cpu") from the driver-core.current tree
and commit fad12ac8c8c2 ("CPU: Introduce ARCH_HAS_CPU_AUTOPROBE and X86
parts") from the driver-core tree.

Just context changes.  I fixed it up (see below) and can carry the fix as
necessary.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc drivers/base/cpu.c
index 23f2c4c,2a0c670..0000000
--- a/drivers/base/cpu.c
+++ b/drivers/base/cpu.c
@@@ -242,7 -224,9 +243,10 @@@ int __cpuinit register_cpu(struct cpu *
  	cpu->node_id = cpu_to_node(num);
  	cpu->dev.id = num;
  	cpu->dev.bus = &cpu_subsys;
 +	cpu->dev.release = cpu_device_release;
+ #ifdef CONFIG_ARCH_HAS_CPU_AUTOPROBE
+ 	cpu->dev.bus->uevent = arch_cpu_uevent;
+ #endif
  	error = device_register(&cpu->dev);
  	if (!error && cpu->hotpluggable)
  		register_cpu_control(cpu);

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Fri, Feb 03, 2012 at 02:51:06PM +1100, Stephen Rothwell wrote:
> Hi Greg,
> 
> Today's linux-next merge of the driver-core tree got a conflict in
> drivers/base/cpu.c between commit 2885e25c422f ("driver core: cpu: remove
> kernel warning when removing a cpu") from the driver-core.current tree
> and commit fad12ac8c8c2 ("CPU: Introduce ARCH_HAS_CPU_AUTOPROBE and X86
> parts") from the driver-core tree.
> 
> Just context changes.  I fixed it up (see below) and can carry the fix as
> necessary.

Thanks for this, I'll fix it up when I merge them back together when
3.3-rc3 comes out.

greg k-h

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Thu, Apr 30, 2020 at 02:25:06PM +1000, Stephen Rothwell wrote:
> Hi all,
> 
> Today's linux-next merge of the driver-core tree got a conflict in:
> 
>   drivers/base/dd.c
> 
> between commits:
> 
>   ce68929f07de ("driver core: Revert default driver_deferred_probe_timeout value to 0")
>   4ccc03e28ec3 ("driver core: Use dev_warn() instead of dev_WARN() for deferred_probe_timeout warnings")
>   35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires")
> 
> from the driver-core.current tree and commit:
> 
>   eb7fbc9fb118 ("driver core: Add missing '\n' in log messages")
> 
> from the driver-core tree.
> 
> I fixed it up (see below) and can carry the fix as necessary. This
> is now fixed as far as linux-next is concerned, but any non trivial
> conflicts should be mentioned to your upstream maintainer when your tree
> is submitted for merging.  You may also want to consider cooperating
> with the maintainer of the conflicting tree to minimise any particularly
> complex conflicts.
> 
> -- 
> Cheers,
> Stephen Rothwell
> 
> diff --cc drivers/base/dd.c
> index 94037be7f5d7,efe6df5bff26..000000000000
> --- a/drivers/base/dd.c
> +++ b/drivers/base/dd.c
> @@@ -258,8 -266,8 +258,8 @@@ int driver_deferred_probe_check_state(s
>   		return -ENODEV;
>   	}
>   
>  -	if (!driver_deferred_probe_timeout) {
>  -		dev_WARN(dev, "deferred probe timeout, ignoring dependency\n");
>  +	if (!driver_deferred_probe_timeout && initcalls_done) {
> - 		dev_warn(dev, "deferred probe timeout, ignoring dependency");
> ++		dev_warn(dev, "deferred probe timeout, ignoring dependency\n");
>   		return -ETIMEDOUT;
>   	}
>   
> @@@ -275,8 -283,7 +275,8 @@@ static void deferred_probe_timeout_work
>   	flush_work(&deferred_probe_work);
>   
>   	list_for_each_entry_safe(private, p, &deferred_probe_pending_list, deferred_probe)
> - 		dev_info(private->device, "deferred probe pending");
> + 		dev_info(private->device, "deferred probe pending\n");
>  +	wake_up(&probe_timeout_waitqueue);
>   }
>   static DECLARE_DELAYED_WORK(deferred_probe_timeout_work, deferred_probe_timeout_work_func);
>   

Looks good, I'll handle this when the .linus branch gets merged into
Linus's tree.

thanks,

greg k-h

linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 2017 bytes --]

Hi all,

Today's linux-next merge of the driver-core tree got a conflict in:

  drivers/base/dd.c

between commits:

  ce68929f07de ("driver core: Revert default driver_deferred_probe_timeout value to 0")
  4ccc03e28ec3 ("driver core: Use dev_warn() instead of dev_WARN() for deferred_probe_timeout warnings")
  35a672363ab3 ("driver core: Ensure wait_for_device_probe() waits until the deferred_probe_timeout fires")

from the driver-core.current tree and commit:

  eb7fbc9fb118 ("driver core: Add missing '\n' in log messages")

from the driver-core tree.

I fixed it up (see below) and can carry the fix as necessary. This
is now fixed as far as linux-next is concerned, but any non trivial
conflicts should be mentioned to your upstream maintainer when your tree
is submitted for merging.  You may also want to consider cooperating
with the maintainer of the conflicting tree to minimise any particularly
complex conflicts.

-- 
Cheers,
Stephen Rothwell

diff --cc drivers/base/dd.c
index 94037be7f5d7,efe6df5bff26..000000000000
--- a/drivers/base/dd.c
+++ b/drivers/base/dd.c
@@@ -258,8 -266,8 +258,8 @@@ int driver_deferred_probe_check_state(s
  		return -ENODEV;
  	}
  
 -	if (!driver_deferred_probe_timeout) {
 -		dev_WARN(dev, "deferred probe timeout, ignoring dependency\n");
 +	if (!driver_deferred_probe_timeout && initcalls_done) {
- 		dev_warn(dev, "deferred probe timeout, ignoring dependency");
++		dev_warn(dev, "deferred probe timeout, ignoring dependency\n");
  		return -ETIMEDOUT;
  	}
  
@@@ -275,8 -283,7 +275,8 @@@ static void deferred_probe_timeout_work
  	flush_work(&deferred_probe_work);
  
  	list_for_each_entry_safe(private, p, &deferred_probe_pending_list, deferred_probe)
- 		dev_info(private->device, "deferred probe pending");
+ 		dev_info(private->device, "deferred probe pending\n");
 +	wake_up(&probe_timeout_waitqueue);
  }
  static DECLARE_DELAYED_WORK(deferred_probe_timeout_work, deferred_probe_timeout_work_func);
  

[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 488 bytes --]

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Mon, Dec 09, 2013 at 02:47:12PM +1100, Stephen Rothwell wrote:
> Hi Greg,
> 
> Today's linux-next merge of the driver-core tree got a conflict in
> fs/sysfs/file.c between commit a8b14744429f ("sysfs: give different
> locking key to regular and bin files") from the driver-core.current tree
> and commit 414985ae23c0 ("sysfs, kernfs: move file core code to
> fs/kernfs/file.c") (among others) from the driver-core tree.
> 
> I just dropped the driver-core.current tree commit as I don't see what
> needs to be done after the other changes in the driver-core tree.

Tejun said he would provide a merge fixup as we knew this was going to
be a problem :)

thanks,

greg k-h

linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 565 bytes --]

Hi Greg,

Today's linux-next merge of the driver-core tree got a conflict in
fs/sysfs/file.c between commit a8b14744429f ("sysfs: give different
locking key to regular and bin files") from the driver-core.current tree
and commit 414985ae23c0 ("sysfs, kernfs: move file core code to
fs/kernfs/file.c") (among others) from the driver-core tree.

I just dropped the driver-core.current tree commit as I don't see what
needs to be done after the other changes in the driver-core tree.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 414 bytes --]

Hi Greg,

On Wed, 19 Jun 2013 20:28:12 -0700 Greg KH <greg@kroah.com> wrote:
>
> Thanks, that helped out a lot.  I've now merged this together, and
> pushed it out to my driver-core-next branch, can you verify that I got
> it all correct?

And I have refetched the driver-core tree for today's linux-next so it
includes that merge.
-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Ming Lei]

On Thu, Jun 20, 2013 at 11:28 AM, Greg KH <greg@kroah.com> wrote:
> On Thu, Jun 20, 2013 at 09:22:13AM +0800, Ming Lei wrote:
>> Hi Stephen,
>>
>> Thanks for your help.
>>
>> On Thu, Jun 20, 2013 at 9:06 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>
>> >
>> > Try "git diff-tree --cc e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d"
>>
>> Greg, Stephen, so is sort of below generated patch what you expected?
>>
>> If yes, I will send it out to you.
>
> Thanks, that helped out a lot.  I've now merged this together, and
> pushed it out to my driver-core-next branch, can you verify that I got
> it all correct?

I just pull your driver-core-next branch, looks drivers/base/firmware_class.c
is fine now, thank you

Thanks,
--
Ming Lei

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Thu, Jun 20, 2013 at 09:22:13AM +0800, Ming Lei wrote:
> Hi Stephen,
> 
> Thanks for your help.
> 
> On Thu, Jun 20, 2013 at 9:06 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> 
> >
> > Try "git diff-tree --cc e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d"
> 
> Greg, Stephen, so is sort of below generated patch what you expected?
> 
> If yes, I will send it out to you.

Thanks, that helped out a lot.  I've now merged this together, and
pushed it out to my driver-core-next branch, can you verify that I got
it all correct?

thanks,

greg k-h

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Ming Lei]

Hi Stephen,

Thanks for your help.

On Thu, Jun 20, 2013 at 9:06 AM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:

>
> Try "git diff-tree --cc e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d"

Greg, Stephen, so is sort of below generated patch what you expected?

If yes, I will send it out to you.

e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d
diff --cc drivers/base/firmware_class.c
index 6ede229,01e2103..7fefcb5
--- a/drivers/base/firmware_class.c
+++ b/drivers/base/firmware_class.c
@@@ -452,35 -446,27 +452,52 @@@ static struct firmware_priv *to_firmwar
  	return container_of(dev, struct firmware_priv, dev);
  }

- static void fw_load_abort(struct firmware_buf *buf)
 -static void fw_load_abort(struct firmware_priv *fw_priv)
++static void __fw_load_abort(struct firmware_buf *buf)
  {
 -	struct firmware_buf *buf = fw_priv->buf;
 -
+ 	/*
+ 	 * There is a small window in which user can write to 'loading'
+ 	 * between loading done and disappearance of 'loading'
+ 	 */
+ 	if (test_bit(FW_STATUS_DONE, &buf->status))
+ 		return;
+
 +	list_del_init(&buf->pending_list);
  	set_bit(FW_STATUS_ABORT, &buf->status);
  	complete_all(&buf->completion);
 +}
 +
++static void fw_load_abort(struct firmware_priv *fw_priv)
++{
++	struct firmware_buf *buf = fw_priv->buf;
++
++	__fw_load_abort(buf);
+
+ 	/* avoid user action after loading abort */
+ 	fw_priv->buf = NULL;
+ }
+
  #define is_fw_load_aborted(buf)	\
  	test_bit(FW_STATUS_ABORT, &(buf)->status)

 +static LIST_HEAD(pending_fw_head);
 +
 +/* reboot notifier for avoid deadlock with usermode_lock */
 +static int fw_shutdown_notify(struct notifier_block *unused1,
 +			      unsigned long unused2, void *unused3)
 +{
 +	mutex_lock(&fw_lock);
 +	while (!list_empty(&pending_fw_head))
- 		fw_load_abort(list_first_entry(&pending_fw_head,
++		__fw_load_abort(list_first_entry(&pending_fw_head,
 +					       struct firmware_buf,
 +					       pending_list));
 +	mutex_unlock(&fw_lock);
 +	return NOTIFY_DONE;
 +}
 +
 +static struct notifier_block fw_shutdown_nb = {
 +	.notifier_call = fw_shutdown_notify,
 +};
 +
  static ssize_t firmware_timeout_show(struct class *class,
  				     struct class_attribute *attr,
  				     char *buf)
@@@ -911,23 -895,6 +927,23 @@@ static int fw_load_from_user_helper(str
  	fw_priv->buf = firmware->priv;
  	return _request_firmware_load(fw_priv, uevent, timeout);
  }
 +
 +#ifdef CONFIG_PM_SLEEP
 +/* kill pending requests without uevent to avoid blocking suspend */
 +static void kill_requests_without_uevent(void)
 +{
 +	struct firmware_buf *buf;
 +	struct firmware_buf *next;
 +
 +	mutex_lock(&fw_lock);
 +	list_for_each_entry_safe(buf, next, &pending_fw_head, pending_list) {
 +		if (!buf->need_uevent)
- 			 fw_load_abort(buf);
++			 __fw_load_abort(buf);
 +	}
 +	mutex_unlock(&fw_lock);
 +}
 +#endif
 +
  #else /* CONFIG_FW_LOADER_USER_HELPER */
  static inline int
  fw_load_from_user_helper(struct firmware *firmware, const char *name,


Thanks,
--
Ming Lei

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 784 bytes --]

Hi Ming,

On Thu, 20 Jun 2013 08:23:30 +0800 Ming Lei <ming.lei@canonical.com> wrote:
>
> Yes, I already merged the two branches together in my local repository,
> and there is one merge commit, but I don't know how to generate/format
> patch for this merge commit only:
> 
> commit e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d
> Merge: bb07b00 8759793
> Author: Ming Lei <tom.leiming@gmail.com>
> Date:   Wed Jun 19 23:56:05 2013 +0800
> 
>     firmware loader: fix conflict driver-core-linus and driver-core-next
> 
>     Conflicts:
>         drivers/base/firmware_class.c
> 
> Anyone who knows, please let me know, thanks.

Try "git diff-tree --cc e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d"

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Ming Lei]

On Thu, Jun 20, 2013 at 12:36 AM, Greg KH <greg@kroah.com> wrote:
>
> Can you merge the two branches together (driver-core-next and
> driver-core-linus) and send me the proper merge patch that I should be
> applying when doing that?  Then I can push that out through the

Yes, I already merged the two branches together in my local repository,
and there is one merge commit, but I don't know how to generate/format
patch for this merge commit only:

commit e4b00d75ee3ed3af9fac83970d21e27d1ad4aa8d
Merge: bb07b00 8759793
Author: Ming Lei <tom.leiming@gmail.com>
Date:   Wed Jun 19 23:56:05 2013 +0800

    firmware loader: fix conflict driver-core-linus and driver-core-next

    Conflicts:
        drivers/base/firmware_class.c

Anyone who knows, please let me know, thanks.

Thanks,
--
Ming Lei

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Wed, Jun 19, 2013 at 11:04:04PM +0800, Ming Lei wrote:
> On Wed, Jun 19, 2013 at 10:39 PM, Greg KH <greg@kroah.com> wrote:
> > On Wed, Jun 19, 2013 at 02:58:39PM +0800, Ming Lei wrote:
> >> On Wed, Jun 19, 2013 at 1:32 PM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> >> > Hi Greg,
> >> >
> >> > Today's linux-next merge of the driver-core tree got a conflict in
> >> > drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
> >> > loader: fix use-after-free by double abort") from the driver-core.current
> >> > tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
> >> > lock at shutdown") from the driver-core tree.
> >> >
> >> > I fixed it up (more may be required - see below) and can carry the fix as
> >> > necessary (no action is required).
> >> >
> >> > --
> >> > Cheers,
> >> > Stephen Rothwell                    sfr@canb.auug.org.au
> >> >
> >> > diff --cc drivers/base/firmware_class.c
> >> > index 01e2103,6ede229..0000000
> >> > --- a/drivers/base/firmware_class.c
> >> > +++ b/drivers/base/firmware_class.c
> >> > @@@ -446,22 -452,11 +452,18 @@@ static struct firmware_priv *to_firmwar
> >> >         return container_of(dev, struct firmware_priv, dev);
> >> >   }
> >> >
> >> > - static void fw_load_abort(struct firmware_priv *fw_priv)
> >> > + static void fw_load_abort(struct firmware_buf *buf)
> >> >   {
> >> > -       struct firmware_buf *buf = fw_priv->buf;
> >> > -
> >> >  +      /*
> >> >  +       * There is a small window in which user can write to 'loading'
> >> >  +       * between loading done and disappearance of 'loading'
> >> >  +       */
> >> >  +      if (test_bit(FW_STATUS_DONE, &buf->status))
> >> >  +              return;
> >> >  +
> >> > +       list_del_init(&buf->pending_list);
> >> >         set_bit(FW_STATUS_ABORT, &buf->status);
> >> >         complete_all(&buf->completion);
> >> > -
> >> > -       /* avoid user action after loading abort */
> >> > -       fw_priv->buf = NULL;
> >>
> >> Hmm, maybe the most important part in the commit 875979368eb4
> >> ("firmware loader: fix use-after-free by double abort") has been removed, :-)
> >>
> >> In fact, the commit 87597936 is for linus tree only because it is a fix,
> >> so the conflict is caused by merging it with other firmware loader patches
> >> in -next tree.
> >>
> >> Greg, I can figure out one patch for -next easily, but it depends you
> >> push it on 3.10-rc or 3.11-rc.
> >
> > I'll be pushing your patch for 3.10-final to Linus as it fixes a bug,
> > but I will need something to resolve the merge issue properly.  Can you
> > provide me that patch/merge?
> 
> OK, I can send you one patch, but I am wondering the patch is against
> today's next tree or your driver-core/driver-core-next?
> 
> If it is against your driver-core/driver-core-next, would you mind letting
> me know how to generate the patch for the conflict?  Sorry for the stupid
> question, because I seldom meet such problem, :-(

Can you merge the two branches together (driver-core-next and
driver-core-linus) and send me the proper merge patch that I should be
applying when doing that?  Then I can push that out through the
driver-core-next tree to make linux-next work properly, as well as make
the merge sane for me when I pull in the final 3.10 release.

thanks,

greg k-h

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Ming Lei]

On Wed, Jun 19, 2013 at 10:39 PM, Greg KH <greg@kroah.com> wrote:
> On Wed, Jun 19, 2013 at 02:58:39PM +0800, Ming Lei wrote:
>> On Wed, Jun 19, 2013 at 1:32 PM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>> > Hi Greg,
>> >
>> > Today's linux-next merge of the driver-core tree got a conflict in
>> > drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
>> > loader: fix use-after-free by double abort") from the driver-core.current
>> > tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
>> > lock at shutdown") from the driver-core tree.
>> >
>> > I fixed it up (more may be required - see below) and can carry the fix as
>> > necessary (no action is required).
>> >
>> > --
>> > Cheers,
>> > Stephen Rothwell                    sfr@canb.auug.org.au
>> >
>> > diff --cc drivers/base/firmware_class.c
>> > index 01e2103,6ede229..0000000
>> > --- a/drivers/base/firmware_class.c
>> > +++ b/drivers/base/firmware_class.c
>> > @@@ -446,22 -452,11 +452,18 @@@ static struct firmware_priv *to_firmwar
>> >         return container_of(dev, struct firmware_priv, dev);
>> >   }
>> >
>> > - static void fw_load_abort(struct firmware_priv *fw_priv)
>> > + static void fw_load_abort(struct firmware_buf *buf)
>> >   {
>> > -       struct firmware_buf *buf = fw_priv->buf;
>> > -
>> >  +      /*
>> >  +       * There is a small window in which user can write to 'loading'
>> >  +       * between loading done and disappearance of 'loading'
>> >  +       */
>> >  +      if (test_bit(FW_STATUS_DONE, &buf->status))
>> >  +              return;
>> >  +
>> > +       list_del_init(&buf->pending_list);
>> >         set_bit(FW_STATUS_ABORT, &buf->status);
>> >         complete_all(&buf->completion);
>> > -
>> > -       /* avoid user action after loading abort */
>> > -       fw_priv->buf = NULL;
>>
>> Hmm, maybe the most important part in the commit 875979368eb4
>> ("firmware loader: fix use-after-free by double abort") has been removed, :-)
>>
>> In fact, the commit 87597936 is for linus tree only because it is a fix,
>> so the conflict is caused by merging it with other firmware loader patches
>> in -next tree.
>>
>> Greg, I can figure out one patch for -next easily, but it depends you
>> push it on 3.10-rc or 3.11-rc.
>
> I'll be pushing your patch for 3.10-final to Linus as it fixes a bug,
> but I will need something to resolve the merge issue properly.  Can you
> provide me that patch/merge?

OK, I can send you one patch, but I am wondering the patch is against
today's next tree or your driver-core/driver-core-next?

If it is against your driver-core/driver-core-next, would you mind letting
me know how to generate the patch for the conflict?  Sorry for the stupid
question, because I seldom meet such problem, :-(


Thanks,
--
Ming Lei

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Wed, Jun 19, 2013 at 02:58:39PM +0800, Ming Lei wrote:
> On Wed, Jun 19, 2013 at 1:32 PM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> > Hi Greg,
> >
> > Today's linux-next merge of the driver-core tree got a conflict in
> > drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
> > loader: fix use-after-free by double abort") from the driver-core.current
> > tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
> > lock at shutdown") from the driver-core tree.
> >
> > I fixed it up (more may be required - see below) and can carry the fix as
> > necessary (no action is required).
> >
> > --
> > Cheers,
> > Stephen Rothwell                    sfr@canb.auug.org.au
> >
> > diff --cc drivers/base/firmware_class.c
> > index 01e2103,6ede229..0000000
> > --- a/drivers/base/firmware_class.c
> > +++ b/drivers/base/firmware_class.c
> > @@@ -446,22 -452,11 +452,18 @@@ static struct firmware_priv *to_firmwar
> >         return container_of(dev, struct firmware_priv, dev);
> >   }
> >
> > - static void fw_load_abort(struct firmware_priv *fw_priv)
> > + static void fw_load_abort(struct firmware_buf *buf)
> >   {
> > -       struct firmware_buf *buf = fw_priv->buf;
> > -
> >  +      /*
> >  +       * There is a small window in which user can write to 'loading'
> >  +       * between loading done and disappearance of 'loading'
> >  +       */
> >  +      if (test_bit(FW_STATUS_DONE, &buf->status))
> >  +              return;
> >  +
> > +       list_del_init(&buf->pending_list);
> >         set_bit(FW_STATUS_ABORT, &buf->status);
> >         complete_all(&buf->completion);
> > -
> > -       /* avoid user action after loading abort */
> > -       fw_priv->buf = NULL;
> 
> Hmm, maybe the most important part in the commit 875979368eb4
> ("firmware loader: fix use-after-free by double abort") has been removed, :-)
> 
> In fact, the commit 87597936 is for linus tree only because it is a fix,
> so the conflict is caused by merging it with other firmware loader patches
> in -next tree.
> 
> Greg, I can figure out one patch for -next easily, but it depends you
> push it on 3.10-rc or 3.11-rc.

I'll be pushing your patch for 3.10-final to Linus as it fixes a bug,
but I will need something to resolve the merge issue properly.  Can you
provide me that patch/merge?

thanks,

greg k-h

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Ming Lei]

On Wed, Jun 19, 2013 at 3:12 PM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> Hi Greg,
>
> On Wed, 19 Jun 2013 15:32:25 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>>
>> Today's linux-next merge of the driver-core tree got a conflict in
>> drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
>> loader: fix use-after-free by double abort") from the driver-core.current
>> tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
>> lock at shutdown") from the driver-core tree.
>>
>> I fixed it up (more may be required - see below) and can carry the fix as
>> necessary (no action is required).
>
> I missed the bit below (at least) and have now added this as a merge fix
> for the driver-core tree.

I think it is OK to merge, then I can add the 'more' part about the original
fix(875979368eb4) against today's change.

>
> From: Stephen Rothwell <sfr@canb.auug.org.au>
> Date: Wed, 19 Jun 2013 17:08:01 +1000
> Subject: [PATCH] fix up for fw_load_abort API change
>
> Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
> ---
>  drivers/base/firmware_class.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c
> index 57aa842..fdacd77 100644
> --- a/drivers/base/firmware_class.c
> +++ b/drivers/base/firmware_class.c
> @@ -812,7 +812,7 @@ static void firmware_class_timeout_work(struct work_struct *work)
>                         struct firmware_priv, timeout_work.work);
>
>         mutex_lock(&fw_lock);
> -       fw_load_abort(fw_priv);
> +       fw_load_abort(fw_priv->buf);
>         mutex_unlock(&fw_lock);
>  }
>
> --
> 1.8.1
>
> --
> Cheers,
> Stephen Rothwell                    sfr@canb.auug.org.au

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1481 bytes --]

Hi Greg,

On Wed, 19 Jun 2013 15:32:25 +1000 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> Today's linux-next merge of the driver-core tree got a conflict in
> drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
> loader: fix use-after-free by double abort") from the driver-core.current
> tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
> lock at shutdown") from the driver-core tree.
> 
> I fixed it up (more may be required - see below) and can carry the fix as
> necessary (no action is required).

I missed the bit below (at least) and have now added this as a merge fix
for the driver-core tree.

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Wed, 19 Jun 2013 17:08:01 +1000
Subject: [PATCH] fix up for fw_load_abort API change

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 drivers/base/firmware_class.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/base/firmware_class.c b/drivers/base/firmware_class.c
index 57aa842..fdacd77 100644
--- a/drivers/base/firmware_class.c
+++ b/drivers/base/firmware_class.c
@@ -812,7 +812,7 @@ static void firmware_class_timeout_work(struct work_struct *work)
 			struct firmware_priv, timeout_work.work);
 
 	mutex_lock(&fw_lock);
-	fw_load_abort(fw_priv);
+	fw_load_abort(fw_priv->buf);
 	mutex_unlock(&fw_lock);
 }
 
-- 
1.8.1

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Ming Lei]

On Wed, Jun 19, 2013 at 1:32 PM, Stephen Rothwell <sfr@canb.auug.org.au> wrote:
> Hi Greg,
>
> Today's linux-next merge of the driver-core tree got a conflict in
> drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
> loader: fix use-after-free by double abort") from the driver-core.current
> tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
> lock at shutdown") from the driver-core tree.
>
> I fixed it up (more may be required - see below) and can carry the fix as
> necessary (no action is required).
>
> --
> Cheers,
> Stephen Rothwell                    sfr@canb.auug.org.au
>
> diff --cc drivers/base/firmware_class.c
> index 01e2103,6ede229..0000000
> --- a/drivers/base/firmware_class.c
> +++ b/drivers/base/firmware_class.c
> @@@ -446,22 -452,11 +452,18 @@@ static struct firmware_priv *to_firmwar
>         return container_of(dev, struct firmware_priv, dev);
>   }
>
> - static void fw_load_abort(struct firmware_priv *fw_priv)
> + static void fw_load_abort(struct firmware_buf *buf)
>   {
> -       struct firmware_buf *buf = fw_priv->buf;
> -
>  +      /*
>  +       * There is a small window in which user can write to 'loading'
>  +       * between loading done and disappearance of 'loading'
>  +       */
>  +      if (test_bit(FW_STATUS_DONE, &buf->status))
>  +              return;
>  +
> +       list_del_init(&buf->pending_list);
>         set_bit(FW_STATUS_ABORT, &buf->status);
>         complete_all(&buf->completion);
> -
> -       /* avoid user action after loading abort */
> -       fw_priv->buf = NULL;

Hmm, maybe the most important part in the commit 875979368eb4
("firmware loader: fix use-after-free by double abort") has been removed, :-)

In fact, the commit 87597936 is for linus tree only because it is a fix,
so the conflict is caused by merging it with other firmware loader patches
in -next tree.

Greg, I can figure out one patch for -next easily, but it depends you
push it on 3.10-rc or 3.11-rc.


Thanks,
--
Ming Lei

Re: linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Greg KH]

On Wed, Jun 19, 2013 at 03:32:25PM +1000, Stephen Rothwell wrote:
> Hi Greg,
> 
> Today's linux-next merge of the driver-core tree got a conflict in
> drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
> loader: fix use-after-free by double abort") from the driver-core.current
> tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
> lock at shutdown") from the driver-core tree.
> 
> I fixed it up (more may be required - see below) and can carry the fix as
> necessary (no action is required).

Thanks, the merge looks correct, Ming, any objection?

greg k-h

linux-next: manual merge of the driver-core tree with the driver-core.current tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 1440 bytes --]

Hi Greg,

Today's linux-next merge of the driver-core tree got a conflict in
drivers/base/firmware_class.c between commit 875979368eb4 ("firmware
loader: fix use-after-free by double abort") from the driver-core.current
tree and commit fe304143b0c3 ("firmware: Avoid deadlock of usermodehelper
lock at shutdown") from the driver-core tree.

I fixed it up (more may be required - see below) and can carry the fix as
necessary (no action is required).

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

diff --cc drivers/base/firmware_class.c
index 01e2103,6ede229..0000000
--- a/drivers/base/firmware_class.c
+++ b/drivers/base/firmware_class.c
@@@ -446,22 -452,11 +452,18 @@@ static struct firmware_priv *to_firmwar
  	return container_of(dev, struct firmware_priv, dev);
  }
  
- static void fw_load_abort(struct firmware_priv *fw_priv)
+ static void fw_load_abort(struct firmware_buf *buf)
  {
- 	struct firmware_buf *buf = fw_priv->buf;
- 
 +	/*
 +	 * There is a small window in which user can write to 'loading'
 +	 * between loading done and disappearance of 'loading'
 +	 */
 +	if (test_bit(FW_STATUS_DONE, &buf->status))
 +		return;
 +
+ 	list_del_init(&buf->pending_list);
  	set_bit(FW_STATUS_ABORT, &buf->status);
  	complete_all(&buf->completion);
- 
- 	/* avoid user action after loading abort */
- 	fw_priv->buf = NULL;
  }
  
  #define is_fw_load_aborted(buf)	\

[-- Attachment #2: Type: application/pgp-signature, Size: 836 bytes --]