Hello Andrew,
On Wed, 26 Feb 2020 at 07:25, Andrew Morton akpm@linux-foundation.org wrote:
The patch titled Subject: dma-buf: free dmabuf->name in dma_buf_release() has been added to the -mm tree. Its filename is dma-buf-free-dmabuf-name-in-dma_buf_release.patch
Thanks for taking this patch via -mm during my absence (I'm just returning from a bit of an illness). If there are other dma-buf patches on your radar that you'd like to take via the mm tree, please let me know and I can provide the necessary Acks. Else I will take them in via drm-misc as usual.
This patch should soon appear at http://ozlabs.org/~akpm/mmots/broken-out/dma-buf-free-dmabuf-name-in-dma_buf... and later at http://ozlabs.org/~akpm/mmotm/broken-out/dma-buf-free-dmabuf-name-in-dma_buf...
Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated there every 3-4 working days
From: Cong Wang xiyou.wangcong@gmail.com Subject: dma-buf: free dmabuf->name in dma_buf_release()
dma-buff name can be set via DMA_BUF_SET_NAME ioctl, but once set it never gets freed.
Free it in dma_buf_release().
Link: http://lkml.kernel.org/r/20200225204446.11378-1-xiyou.wangcong@gmail.com Fixes: bb2bb9030425 ("dma-buf: add DMA_BUF_SET_NAME ioctls") Reported-by: syzbot+b2098bc44728a4efb3e9@syzkaller.appspotmail.com Signed-off-by: Cong Wang xiyou.wangcong@gmail.com Reviewed-by: Andrew Morton akpm@linux-foundation.org Acked-by: Chenbo Feng fengc@google.com Cc: Sumit Semwal sumit.semwal@linaro.org Cc: Greg Hackmann ghackmann@google.com Signed-off-by: Andrew Morton akpm@linux-foundation.org
drivers/dma-buf/dma-buf.c | 1 + 1 file changed, 1 insertion(+)
--- a/drivers/dma-buf/dma-buf.c~dma-buf-free-dmabuf-name-in-dma_buf_release +++ a/drivers/dma-buf/dma-buf.c @@ -108,6 +108,7 @@ static int dma_buf_release(struct inode dma_resv_fini(dmabuf->resv);
module_put(dmabuf->owner);
kfree(dmabuf->name); kfree(dmabuf); return 0;} _
Patches currently in -mm which might be from xiyou.wangcong@gmail.com are
dma-buf-free-dmabuf-name-in-dma_buf_release.patch
Thanks and best regards, Sumit.
On Wed, Feb 26, 2020 at 5:29 AM Sumit Semwal sumit.semwal@linaro.org wrote:
Hello Andrew,
On Wed, 26 Feb 2020 at 07:25, Andrew Morton akpm@linux-foundation.org wrote:
The patch titled Subject: dma-buf: free dmabuf->name in dma_buf_release() has been added to the -mm tree. Its filename is dma-buf-free-dmabuf-name-in-dma_buf_release.patch
Thanks for taking this patch via -mm during my absence (I'm just returning from a bit of an illness). If there are other dma-buf patches on your radar that you'd like to take via the mm tree, please let me know and I can provide the necessary Acks. Else I will take them in via drm-misc as usual.
I thought at least that for cases like these -mm is the last resort tree, so proper thing to do here is apply this fix to drm-misc-fixes and get it out there. -mm rebases, so will fall out again.
Also maybe we should have had some tests for this, plus some integration with the gem object name instead of duplicating buffer object names everywhere for different reasons. -Daniel
This patch should soon appear at http://ozlabs.org/~akpm/mmots/broken-out/dma-buf-free-dmabuf-name-in-dma_buf... and later at http://ozlabs.org/~akpm/mmotm/broken-out/dma-buf-free-dmabuf-name-in-dma_buf...
Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/process/submit-checklist.rst when testing your code ***
The -mm tree is included into linux-next and is updated there every 3-4 working days
From: Cong Wang xiyou.wangcong@gmail.com Subject: dma-buf: free dmabuf->name in dma_buf_release()
dma-buff name can be set via DMA_BUF_SET_NAME ioctl, but once set it never gets freed.
Free it in dma_buf_release().
Link: http://lkml.kernel.org/r/20200225204446.11378-1-xiyou.wangcong@gmail.com Fixes: bb2bb9030425 ("dma-buf: add DMA_BUF_SET_NAME ioctls") Reported-by: syzbot+b2098bc44728a4efb3e9@syzkaller.appspotmail.com Signed-off-by: Cong Wang xiyou.wangcong@gmail.com Reviewed-by: Andrew Morton akpm@linux-foundation.org Acked-by: Chenbo Feng fengc@google.com Cc: Sumit Semwal sumit.semwal@linaro.org Cc: Greg Hackmann ghackmann@google.com Signed-off-by: Andrew Morton akpm@linux-foundation.org
drivers/dma-buf/dma-buf.c | 1 + 1 file changed, 1 insertion(+)
--- a/drivers/dma-buf/dma-buf.c~dma-buf-free-dmabuf-name-in-dma_buf_release +++ a/drivers/dma-buf/dma-buf.c @@ -108,6 +108,7 @@ static int dma_buf_release(struct inode dma_resv_fini(dmabuf->resv);
module_put(dmabuf->owner);
kfree(dmabuf->name); kfree(dmabuf); return 0;} _
Patches currently in -mm which might be from xiyou.wangcong@gmail.com are
dma-buf-free-dmabuf-name-in-dma_buf_release.patch
Thanks and best regards, Sumit. _______________________________________________ dri-devel mailing list dri-devel@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/dri-devel
On Wed, 26 Feb 2020 10:36:26 +0100 Daniel Vetter daniel@ffwll.ch wrote:
On Wed, Feb 26, 2020 at 5:29 AM Sumit Semwal sumit.semwal@linaro.org wrote:
Hello Andrew,
On Wed, 26 Feb 2020 at 07:25, Andrew Morton akpm@linux-foundation.org wrote:
The patch titled Subject: dma-buf: free dmabuf->name in dma_buf_release() has been added to the -mm tree. Its filename is dma-buf-free-dmabuf-name-in-dma_buf_release.patch
Thanks for taking this patch via -mm during my absence (I'm just returning from a bit of an illness). If there are other dma-buf patches on your radar that you'd like to take via the mm tree, please let me know and I can provide the necessary Acks. Else I will take them in via drm-misc as usual.
I thought at least that for cases like these -mm is the last resort tree, so proper thing to do here is apply this fix to drm-misc-fixes and get it out there. -mm rebases, so will fall out again.
Yup, go ahead. If a patch pops up in linux-next I'll autodrop by copy.
And please do give some thought to whether this should be cc:stable. If it's an unprivileged operation then hellyeah.
dri-devel@lists.freedesktop.org
-
Andrew Morton -
Daniel Vetter -
Sumit Semwal