linux-next: build failure after merge of the staging-next tree

linux-next: build failure after merge of the staging-next tree

[Stephen Rothwell]

Hi Greg,

After merging the staging-next tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/md/md.c: In function 'level_store':
drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'

Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
Implement sysfs tagged directory support") from the driver-core tree
interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
manage redundancy group in sysfs when changing level") from the md.

I have applied this fixup patch for today and can carry it as necessary.

[This could have been avoided, of course, by creating a new API (maybe
sysfs_get_dirent_tagged) and implementing the old API in terms of that].

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Fri, 30 Apr 2010 15:00:20 +1000
Subject: [PATCH] md: fixup for sysfs_get_dirent API change

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 drivers/md/md.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/drivers/md/md.c b/drivers/md/md.c
index 7f23558..b075b26 100644
--- a/drivers/md/md.c
+++ b/drivers/md/md.c
@@ -3026,7 +3026,7 @@ level_store(mddev_t *mddev, const char *buf, size_t len)
 			printk(KERN_WARNING
 			       "md: cannot register extra attributes for %s\n",
 			       mdname(mddev));
-		mddev->sysfs_action = sysfs_get_dirent(mddev->kobj.sd, "sync_action");
+		mddev->sysfs_action = sysfs_get_dirent(mddev->kobj.sd, NULL, "sync_action");
 	}		
 	if (mddev->pers->sync_request != NULL &&
 	    pers->sync_request == NULL) {
-- 
1.7.1

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

Re: linux-next: build failure after merge of the staging-next tree

[Greg KH]

On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
> Hi Greg,
> 
> After merging the staging-next tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> drivers/md/md.c: In function 'level_store':
> drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
> 
> Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
> Implement sysfs tagged directory support") from the driver-core tree
> interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
> manage redundancy group in sysfs when changing level") from the md.
> 
> I have applied this fixup patch for today and can carry it as necessary.
> 
> [This could have been avoided, of course, by creating a new API (maybe
> sysfs_get_dirent_tagged) and implementing the old API in terms of that].

Thanks for the fixup.

greg k-h

Re: linux-next: build failure after merge of the staging-next tree

[Neil Brown]

On Fri, 30 Apr 2010 07:43:40 -0700
Greg KH <greg@kroah.com> wrote:

> On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
> > Hi Greg,
> > 
> > After merging the staging-next tree, today's linux-next build (x86_64
> > allmodconfig) failed like this:
> > 
> > drivers/md/md.c: In function 'level_store':
> > drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
> > 
> > Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
> > Implement sysfs tagged directory support") from the driver-core tree
> > interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
> > manage redundancy group in sysfs when changing level") from the md.
> > 
> > I have applied this fixup patch for today and can carry it as necessary.
> > 
> > [This could have been avoided, of course, by creating a new API (maybe
> > sysfs_get_dirent_tagged) and implementing the old API in terms of that].
> 
> Thanks for the fixup.

hi Greg,
 I'm assuming that you will submit you patch queue to Linus first, then I
 will create a merge which fixes this issue and ask Linus to pull that - so
 everything is all still bisectable without me rebasing.

 If you are OK with that, please let me know when you have submitted your
 queue..  I'll try to keep and eye open, but I might miss it...

Thanks,
NeilBrown

Re: linux-next: build failure after merge of the staging-next tree

[Greg KH]

On Tue, May 18, 2010 at 03:30:20PM +1000, Neil Brown wrote:
> On Fri, 30 Apr 2010 07:43:40 -0700
> Greg KH <greg@kroah.com> wrote:
> 
> > On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
> > > Hi Greg,
> > > 
> > > After merging the staging-next tree, today's linux-next build (x86_64
> > > allmodconfig) failed like this:
> > > 
> > > drivers/md/md.c: In function 'level_store':
> > > drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
> > > 
> > > Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
> > > Implement sysfs tagged directory support") from the driver-core tree
> > > interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
> > > manage redundancy group in sysfs when changing level") from the md.
> > > 
> > > I have applied this fixup patch for today and can carry it as necessary.
> > > 
> > > [This could have been avoided, of course, by creating a new API (maybe
> > > sysfs_get_dirent_tagged) and implementing the old API in terms of that].
> > 
> > Thanks for the fixup.
> 
> hi Greg,
>  I'm assuming that you will submit you patch queue to Linus first, then I
>  will create a merge which fixes this issue and ask Linus to pull that - so
>  everything is all still bisectable without me rebasing.
> 
>  If you are OK with that, please let me know when you have submitted your
>  queue..  I'll try to keep and eye open, but I might miss it...

The drivercore queue is now in Linus's tree.

Oh, you missed a kref_set() usage in the tree as well, you might want to
fix that up also.

thanks,

greg k-h

Re: linux-next: build failure after merge of the staging-next tree

[Neil Brown]

On Fri, 21 May 2010 16:36:55 -0700
Greg KH <greg@kroah.com> wrote:

> On Tue, May 18, 2010 at 03:30:20PM +1000, Neil Brown wrote:
> > On Fri, 30 Apr 2010 07:43:40 -0700
> > Greg KH <greg@kroah.com> wrote:
> > 
> > > On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
> > > > Hi Greg,
> > > > 
> > > > After merging the staging-next tree, today's linux-next build (x86_64
> > > > allmodconfig) failed like this:
> > > > 
> > > > drivers/md/md.c: In function 'level_store':
> > > > drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
> > > > 
> > > > Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
> > > > Implement sysfs tagged directory support") from the driver-core tree
> > > > interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
> > > > manage redundancy group in sysfs when changing level") from the md.
> > > > 
> > > > I have applied this fixup patch for today and can carry it as necessary.
> > > > 
> > > > [This could have been avoided, of course, by creating a new API (maybe
> > > > sysfs_get_dirent_tagged) and implementing the old API in terms of that].
> > > 
> > > Thanks for the fixup.
> > 
> > hi Greg,
> >  I'm assuming that you will submit you patch queue to Linus first, then I
> >  will create a merge which fixes this issue and ask Linus to pull that - so
> >  everything is all still bisectable without me rebasing.
> > 
> >  If you are OK with that, please let me know when you have submitted your
> >  queue..  I'll try to keep and eye open, but I might miss it...
> 
> The drivercore queue is now in Linus's tree.

Thanks Greg.
I've created a merge which resolves the sysfs_get_dirent issue and sent Linus
a pull request.

> 
> Oh, you missed a kref_set() usage in the tree as well, you might want to
> fix that up also.

That would be the one in fs/ceph ??
Sage knows about that and presumably has a fix queued.
  http://lkml.org/lkml/2010/3/17/224

Thanks,
NeilBrown



> 
> thanks,
> 
> greg k-h
> --
> To unsubscribe from this list: send the line "unsubscribe linux-next" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: linux-next: build failure after merge of the staging-next tree

[Greg KH]

On Sat, May 22, 2010 at 10:20:48AM +1000, Neil Brown wrote:
> On Fri, 21 May 2010 16:36:55 -0700
> Greg KH <greg@kroah.com> wrote:
> 
> > On Tue, May 18, 2010 at 03:30:20PM +1000, Neil Brown wrote:
> > > On Fri, 30 Apr 2010 07:43:40 -0700
> > > Greg KH <greg@kroah.com> wrote:
> > > 
> > > > On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
> > > > > Hi Greg,
> > > > > 
> > > > > After merging the staging-next tree, today's linux-next build (x86_64
> > > > > allmodconfig) failed like this:
> > > > > 
> > > > > drivers/md/md.c: In function 'level_store':
> > > > > drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
> > > > > 
> > > > > Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
> > > > > Implement sysfs tagged directory support") from the driver-core tree
> > > > > interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
> > > > > manage redundancy group in sysfs when changing level") from the md.
> > > > > 
> > > > > I have applied this fixup patch for today and can carry it as necessary.
> > > > > 
> > > > > [This could have been avoided, of course, by creating a new API (maybe
> > > > > sysfs_get_dirent_tagged) and implementing the old API in terms of that].
> > > > 
> > > > Thanks for the fixup.
> > > 
> > > hi Greg,
> > >  I'm assuming that you will submit you patch queue to Linus first, then I
> > >  will create a merge which fixes this issue and ask Linus to pull that - so
> > >  everything is all still bisectable without me rebasing.
> > > 
> > >  If you are OK with that, please let me know when you have submitted your
> > >  queue..  I'll try to keep and eye open, but I might miss it...
> > 
> > The drivercore queue is now in Linus's tree.
> 
> Thanks Greg.
> I've created a merge which resolves the sysfs_get_dirent issue and sent Linus
> a pull request.

wonderful.

> > Oh, you missed a kref_set() usage in the tree as well, you might want to
> > fix that up also.
> 
> That would be the one in fs/ceph ??
> Sage knows about that and presumably has a fix queued.
>   http://lkml.org/lkml/2010/3/17/224

Ah, yes, I thought I had remembered something about that.

greg k-h

Re: linux-next: build failure after merge of the staging-next tree

[Sage Weil]

On Sat, 22 May 2010, Greg KH wrote:
> > > Oh, you missed a kref_set() usage in the tree as well, you might want to
> > > fix that up also.
> > 
> > That would be the one in fs/ceph ??
> > Sage knows about that and presumably has a fix queued.
> >   http://lkml.org/lkml/2010/3/17/224
> 
> Ah, yes, I thought I had remembered something about that.

Yep, I just sent off the pull request that includes the kref_set() fix.

sage

Re: linux-next: build failure after merge of the staging-next tree

[Greg KH]

On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
> Hi Greg,
> 
> After merging the staging-next tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> drivers/md/md.c: In function 'level_store':
> drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
> 
> Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
> Implement sysfs tagged directory support") from the driver-core tree
> interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
> manage redundancy group in sysfs when changing level") from the md.
> 
> I have applied this fixup patch for today and can carry it as necessary.

Thanks, that would be great.

> [This could have been avoided, of course, by creating a new API (maybe
> sysfs_get_dirent_tagged) and implementing the old API in terms of that].

Hm Eric, any thoughts?

thanks,

greg k-h

Re: linux-next: build failure after merge of the staging-next tree

[Eric W. Biederman]

Greg KH <greg@kroah.com> writes:

> On Fri, Apr 30, 2010 at 03:52:05PM +1000, Stephen Rothwell wrote:
>> Hi Greg,
>> 
>> After merging the staging-next tree, today's linux-next build (x86_64
>> allmodconfig) failed like this:
>> 
>> drivers/md/md.c: In function 'level_store':
>> drivers/md/md.c:3029: error: too few arguments to function 'sysfs_get_dirent'
>> 
>> Caused by commit 262f8e4937e7b4a587923ca3c039a184668f49ec ("sysfs:
>> Implement sysfs tagged directory support") from the driver-core tree
>> interacting with commit fecc531e3cc0de60514d326c7d82f1075ed55888 ("md:
>> manage redundancy group in sysfs when changing level") from the md.
>> 
>> I have applied this fixup patch for today and can carry it as necessary.
>
> Thanks, that would be great.
>
>> [This could have been avoided, of course, by creating a new API (maybe
>> sysfs_get_dirent_tagged) and implementing the old API in terms of that].
>
> Hm Eric, any thoughts?

I believe I touched all of the users of sysfs_get_dirent outside of sysfs
that existed when I sent you the patch.

Right now using sysfs_get_dirent is a hack to support notifications of
changes to sysfs files from atomic contexts, where sysfs_notify is
unsafe because it sleeps on sysfs_mutex.  There are maybe 5 callers of
sysfs_get_dirent outside of sysfs.  Given that we have a nice compile
time error I don't think it makes sense to have multiple versions of the
function.

It will probably makes sense at some point to go through and push
everything to using a less hacky solution, but for the moment the
solution is correct and doesn't cause too much pain so I'm not too
worried about it.

As for the fixup patch itself it looked correct.

Eric

linux-next: build failure after merge of the staging-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 762 bytes --]

Hi Greg,

After merging the staging-next tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/media/video/tlg2300/pd-main.c: In function 'poseidon_probe':
drivers/media/video/tlg2300/pd-main.c:457: error: 'struct usb_device' has no member named 'autosuspend_disabled'
drivers/net/wimax/i2400m/usb.c: In function 'i2400mu_probe':
drivers/net/wimax/i2400m/usb.c:508: error: 'struct usb_device' has no member named 'autosuspend_disabled'

Caused by commit 3a03bb2443e8fb3b35666dc598e7f626519c544a ("USB: use PM
core routines to enable/disable autosuspend") from the usb tree.  I have
reverted that commit for today.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]

Re: linux-next: build failure after merge of the staging-next tree

[Greg KH]

On Fri, Apr 30, 2010 at 03:52:24PM +1000, Stephen Rothwell wrote:
> Hi Greg,
> 
> After merging the staging-next tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> drivers/media/video/tlg2300/pd-main.c: In function 'poseidon_probe':
> drivers/media/video/tlg2300/pd-main.c:457: error: 'struct usb_device' has no member named 'autosuspend_disabled'
> drivers/net/wimax/i2400m/usb.c: In function 'i2400mu_probe':
> drivers/net/wimax/i2400m/usb.c:508: error: 'struct usb_device' has no member named 'autosuspend_disabled'
> 
> Caused by commit 3a03bb2443e8fb3b35666dc598e7f626519c544a ("USB: use PM
> core routines to enable/disable autosuspend") from the usb tree.  I have
> reverted that commit for today.

It would be good to get that back into the tree.  Alan, care to fix this
up?

thanks,

greg k-h

linux-next: build failure after merge of the staging-next tree

[Stephen Rothwell]

Hi Greg,

After merging the staging-next tree, today's linux-next build (x86_64
allmodconfig) failed like this:

drivers/net/usb/ipheth.c: In function 'ipheth_alloc_urbs':
drivers/net/usb/ipheth.c:131: error: implicit declaration of function 'usb_buffer_alloc'
drivers/net/usb/ipheth.c:134: warning: assignment makes pointer from integer without a cast
drivers/net/usb/ipheth.c:141: warning: assignment makes pointer from integer without a cast
drivers/net/usb/ipheth.c:153: error: implicit declaration of function 'usb_buffer_free'

Caused by commit 3cc96f50329227fbf34de2459491901b6ee1c98c ("USB: rename
usb_buffer_alloc() and usb_buffer_free() users") from the usb tree
interacting with commit a19259c3d589a014e5f47f148f74dfc44422c82b
("drivers/net/usb: Add new driver ipheth") which entered Linus' tree about
April 25 via the net tree.

I applied the following fixup patch for today.

From: Stephen Rothwell <sfr@canb.auug.org.au>
Date: Fri, 30 Apr 2010 15:44:06 +1000
Subject: [PATCH] net: fix for usb_buffer_alloc/free rename

Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au>
---
 drivers/net/usb/ipheth.c |   10 +++++-----
 1 files changed, 5 insertions(+), 5 deletions(-)

diff --git a/drivers/net/usb/ipheth.c b/drivers/net/usb/ipheth.c
index 418825d..41071d0 100644
--- a/drivers/net/usb/ipheth.c
+++ b/drivers/net/usb/ipheth.c
@@ -128,14 +128,14 @@ static int ipheth_alloc_urbs(struct ipheth_device *iphone)
 	if (rx_urb == NULL)
 		goto free_tx_urb;
 
-	tx_buf = usb_buffer_alloc(iphone->udev,
+	tx_buf = usb_alloc_coherent(iphone->udev,
 				  IPHETH_BUF_SIZE,
 				  GFP_KERNEL,
 				  &tx_urb->transfer_dma);
 	if (tx_buf == NULL)
 		goto free_rx_urb;
 
-	rx_buf = usb_buffer_alloc(iphone->udev,
+	rx_buf = usb_alloc_coherent(iphone->udev,
 				  IPHETH_BUF_SIZE,
 				  GFP_KERNEL,
 				  &rx_urb->transfer_dma);
@@ -150,7 +150,7 @@ static int ipheth_alloc_urbs(struct ipheth_device *iphone)
 	return 0;
 
 free_tx_buf:
-	usb_buffer_free(iphone->udev, IPHETH_BUF_SIZE, tx_buf,
+	usb_free_coherent(iphone->udev, IPHETH_BUF_SIZE, tx_buf,
 			tx_urb->transfer_dma);
 free_rx_urb:
 	usb_free_urb(rx_urb);
@@ -162,9 +162,9 @@ error_nomem:
 
 static void ipheth_free_urbs(struct ipheth_device *iphone)
 {
-	usb_buffer_free(iphone->udev, IPHETH_BUF_SIZE, iphone->rx_buf,
+	usb_free_coherent(iphone->udev, IPHETH_BUF_SIZE, iphone->rx_buf,
 			iphone->rx_urb->transfer_dma);
-	usb_buffer_free(iphone->udev, IPHETH_BUF_SIZE, iphone->tx_buf,
+	usb_free_coherent(iphone->udev, IPHETH_BUF_SIZE, iphone->tx_buf,
 			iphone->tx_urb->transfer_dma);
 	usb_free_urb(iphone->rx_urb);
 	usb_free_urb(iphone->tx_urb);
-- 
1.7.1

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

Re: linux-next: build failure after merge of the staging-next tree

[Greg KH]

On Fri, Apr 30, 2010 at 03:50:52PM +1000, Stephen Rothwell wrote:
> Hi Greg,
> 
> After merging the staging-next tree, today's linux-next build (x86_64
> allmodconfig) failed like this:
> 
> drivers/net/usb/ipheth.c: In function 'ipheth_alloc_urbs':
> drivers/net/usb/ipheth.c:131: error: implicit declaration of function 'usb_buffer_alloc'
> drivers/net/usb/ipheth.c:134: warning: assignment makes pointer from integer without a cast
> drivers/net/usb/ipheth.c:141: warning: assignment makes pointer from integer without a cast
> drivers/net/usb/ipheth.c:153: error: implicit declaration of function 'usb_buffer_free'
> 
> Caused by commit 3cc96f50329227fbf34de2459491901b6ee1c98c ("USB: rename
> usb_buffer_alloc() and usb_buffer_free() users") from the usb tree
> interacting with commit a19259c3d589a014e5f47f148f74dfc44422c82b
> ("drivers/net/usb: Add new driver ipheth") which entered Linus' tree about
> April 25 via the net tree.
> 
> I applied the following fixup patch for today.

Thanks, I'll be sending in the patch that adds the api call to Linus
today so that the individual trees can make these changes.

thanks,

greg k-h

Re: linux-next: build failure after merge of the staging-next tree

[Stephen Rothwell]

[-- Attachment #1: Type: text/plain, Size: 357 bytes --]

Hi Greg,

On Fri, 30 Apr 2010 07:45:07 -0700 Greg KH <greg@kroah.com> wrote:
>
> Thanks, I'll be sending in the patch that adds the api call to Linus
> today so that the individual trees can make these changes.

That will be great, thanks.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au
http://www.canb.auug.org.au/~sfr/

[-- Attachment #2: Type: application/pgp-signature, Size: 198 bytes --]