Hey Jens,
I've got two tiny fixes for the xen-blkback (see below):
git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git for-jens/xen-blkback.fixes
and they are based on for-2.6.40/drivers branch. Is that the way you
prefer to pull patches from sub-trees? Or should I rebase it on 3.0-rc1?
Or should send bug-fixes just directly to Linus?
Dan Carpenter (1):
xen/blkback: potential null dereference in error handling
Laszlo Ersek (1):
xen/blkback: don't call vbd_size() if bd_disk is NULL
drivers/block/xen-blkback/blkback.c | 10 ++++++----
drivers/block/xen-blkback/xenbus.c | 3 +--
2 files changed, 7 insertions(+), 6 deletions(-)
And them combined:
diff --git a/drivers/block/xen-blkback/blkback.c b/drivers/block/xen-blkback/blkback.c
index c73910c..e3c26d1 100644
--- a/drivers/block/xen-blkback/blkback.c
+++ b/drivers/block/xen-blkback/blkback.c
@@ -809,11 +809,13 @@ static int __init xen_blkif_init(void)
failed_init:
kfree(blkbk->pending_reqs);
kfree(blkbk->pending_grant_handles);
- for (i = 0; i < mmap_pages; i++) {
- if (blkbk->pending_pages[i])
- __free_page(blkbk->pending_pages[i]);
+ if (blkback->pending_pages) {
+ for (i = 0; i < mmap_pages; i++) {
+ if (blkbk->pending_pages[i])
+ __free_page(blkbk->pending_pages[i]);
+ }
+ kfree(blkbk->pending_pages);
}
- kfree(blkbk->pending_pages);
kfree(blkbk);
blkbk = NULL;
return rc;
diff --git a/drivers/block/xen-blkback/xenbus.c b/drivers/block/xen-blkback/xenbus.c
index 3457082..6cc0db1 100644
--- a/drivers/block/xen-blkback/xenbus.c
+++ b/drivers/block/xen-blkback/xenbus.c
@@ -357,14 +357,13 @@ static int xen_vbd_create(struct xen_blkif *blkif, blkif_vdev_t handle,
}
vbd->bdev = bdev;
- vbd->size = vbd_sz(vbd);
-
if (vbd->bdev->bd_disk == NULL) {
DPRINTK("xen_vbd_create: device %08x doesn't exist.\n",
vbd->pdevice);
xen_vbd_free(vbd);
return -ENOENT;
}
+ vbd->size = vbd_sz(vbd);
if (vbd->bdev->bd_disk->flags & GENHD_FL_CD || cdrom)
vbd->type |= VDISK_CDROM;
On 2011-05-31 21:10, Konrad Rzeszutek Wilk wrote:
> Hey Jens,
>
> I've got two tiny fixes for the xen-blkback (see below):
>
> git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git for-jens/xen-blkback.fixes
>
> and they are based on for-2.6.40/drivers branch. Is that the way you
> prefer to pull patches from sub-trees? Or should I rebase it on 3.0-rc1?
> Or should send bug-fixes just directly to Linus?
Once I send off the big initial merges which are in the version named
for-2.6.40/xx branches, anything after that should just be based on my
for-linus branch. That's what I use for post-rc1 merges with Linus.
But the only real sticking point is that your branch should not contain
something that I don't have in my for-linus. And you don't right now, so
I just pulled it :-)
--
Jens Axboe
On Tue, May 31, 2011 at 09:25:11PM +0200, Jens Axboe wrote:
> On 2011-05-31 21:10, Konrad Rzeszutek Wilk wrote:
> > Hey Jens,
> >
> > I've got two tiny fixes for the xen-blkback (see below):
> >
> > git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git for-jens/xen-blkback.fixes
> >
> > and they are based on for-2.6.40/drivers branch. Is that the way you
> > prefer to pull patches from sub-trees? Or should I rebase it on 3.0-rc1?
> > Or should send bug-fixes just directly to Linus?
>
> Once I send off the big initial merges which are in the version named
> for-2.6.40/xx branches, anything after that should just be based on my
> for-linus branch. That's what I use for post-rc1 merges with Linus.
>
> But the only real sticking point is that your branch should not contain
> something that I don't have in my for-linus. And you don't right now, so
> I just pulled it :-)
Can you repull that branch please? I've rebased it on for-linus branch
and fixed the embarrassing compilation issue it had..
On 2011-06-01 17:15, Konrad Rzeszutek Wilk wrote:
> On Tue, May 31, 2011 at 09:25:11PM +0200, Jens Axboe wrote:
>> On 2011-05-31 21:10, Konrad Rzeszutek Wilk wrote:
>>> Hey Jens,
>>>
>>> I've got two tiny fixes for the xen-blkback (see below):
>>>
>>> git://git.kernel.org/pub/scm/linux/kernel/git/konrad/xen.git for-jens/xen-blkback.fixes
>>>
>>> and they are based on for-2.6.40/drivers branch. Is that the way you
>>> prefer to pull patches from sub-trees? Or should I rebase it on 3.0-rc1?
>>> Or should send bug-fixes just directly to Linus?
>>
>> Once I send off the big initial merges which are in the version named
>> for-2.6.40/xx branches, anything after that should just be based on my
>> for-linus branch. That's what I use for post-rc1 merges with Linus.
>>
>> But the only real sticking point is that your branch should not contain
>> something that I don't have in my for-linus. And you don't right now, so
>> I just pulled it :-)
>
> Can you repull that branch please? I've rebased it on for-linus branch
> and fixed the embarrassing compilation issue it had..
Repulled, and it seems to compile as well this time :-)
--
Jens Axboe