* [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb @ 2021-01-11 10:49 Bui Quang Minh 2021-01-11 12:00 ` Oliver Neukum 0 siblings, 1 reply; 5+ messages in thread From: Bui Quang Minh @ 2021-01-11 10:49 UTC (permalink / raw) To: linux-usb Cc: a.darwish, bigeasy, gregkh, linux-kernel, minhquangbui99, stern, syzkaller-bugs, tglx In mcba_usb_read_bulk_callback(), when we don't resubmit or fails to resubmit the urb, we need to deallocate the transfer buffer that is allocated in mcba_usb_start(). Reported-by: syzbot+57281c762a3922e14dfe@syzkaller.appspotmail.com Signed-off-by: Bui Quang Minh <minhquangbui99@gmail.com> --- v1: add memory leak fix when not resubmitting urb v2: add memory leak fix when failing to resubmit urb drivers/net/can/usb/mcba_usb.c | 11 ++++++++--- 1 file changed, 8 insertions(+), 3 deletions(-) diff --git a/drivers/net/can/usb/mcba_usb.c b/drivers/net/can/usb/mcba_usb.c index df54eb7d4b36..30236e640116 100644 --- a/drivers/net/can/usb/mcba_usb.c +++ b/drivers/net/can/usb/mcba_usb.c @@ -584,6 +584,8 @@ static void mcba_usb_read_bulk_callback(struct urb *urb) case -EPIPE: case -EPROTO: case -ESHUTDOWN: + usb_free_coherent(urb->dev, urb->transfer_buffer_length, + urb->transfer_buffer, urb->transfer_dma); return; default: @@ -615,11 +617,14 @@ static void mcba_usb_read_bulk_callback(struct urb *urb) retval = usb_submit_urb(urb, GFP_ATOMIC); - if (retval == -ENODEV) - netif_device_detach(netdev); - else if (retval) + if (retval < 0) { netdev_err(netdev, "failed resubmitting read bulk urb: %d\n", retval); + usb_free_coherent(urb->dev, urb->transfer_buffer_length, + urb->transfer_buffer, urb->transfer_dma); + if (retval == -ENODEV) + netif_device_detach(netdev); + } } /* Start USB device */ -- 2.17.1 ^ permalink raw reply related [flat|nested] 5+ messages in thread
* Re: [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb 2021-01-11 10:49 [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb Bui Quang Minh @ 2021-01-11 12:00 ` Oliver Neukum 2021-01-11 14:31 ` Bui Quang Minh 0 siblings, 1 reply; 5+ messages in thread From: Oliver Neukum @ 2021-01-11 12:00 UTC (permalink / raw) To: Bui Quang Minh, linux-usb Cc: a.darwish, bigeasy, gregkh, linux-kernel, stern, syzkaller-bugs, tglx Am Montag, den 11.01.2021, 10:49 +0000 schrieb Bui Quang Minh: > In mcba_usb_read_bulk_callback(), when we don't resubmit or fails to > resubmit the urb, we need to deallocate the transfer buffer that is > allocated in mcba_usb_start(). > > Reported-by: syzbot+57281c762a3922e14dfe@syzkaller.appspotmail.com > Signed-off-by: Bui Quang Minh <minhquangbui99@gmail.com> > --- > v1: add memory leak fix when not resubmitting urb > v2: add memory leak fix when failing to resubmit urb > > drivers/net/can/usb/mcba_usb.c | 11 ++++++++--- > 1 file changed, 8 insertions(+), 3 deletions(-) > > diff --git a/drivers/net/can/usb/mcba_usb.c b/drivers/net/can/usb/mcba_usb.c > index df54eb7d4b36..30236e640116 100644 > --- a/drivers/net/can/usb/mcba_usb.c > +++ b/drivers/net/can/usb/mcba_usb.c > @@ -584,6 +584,8 @@ static void mcba_usb_read_bulk_callback(struct urb *urb) > case -EPIPE: > case -EPROTO: > case -ESHUTDOWN: > + usb_free_coherent(urb->dev, urb->transfer_buffer_length, > + urb->transfer_buffer, urb->transfer_dma); > return; > Can you call usb_free_coherent() in what can be hard IRQ context? Regards Oliver ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb 2021-01-11 12:00 ` Oliver Neukum @ 2021-01-11 14:31 ` Bui Quang Minh 2021-01-12 6:42 ` Minh Bùi Quang 0 siblings, 1 reply; 5+ messages in thread From: Bui Quang Minh @ 2021-01-11 14:31 UTC (permalink / raw) To: Oliver Neukum Cc: linux-usb, a.darwish, bigeasy, gregkh, linux-kernel, stern, syzkaller-bugs, tglx On Mon, Jan 11, 2021 at 01:00:31PM +0100, Oliver Neukum wrote: > Am Montag, den 11.01.2021, 10:49 +0000 schrieb Bui Quang Minh: > > In mcba_usb_read_bulk_callback(), when we don't resubmit or fails to > > resubmit the urb, we need to deallocate the transfer buffer that is > > allocated in mcba_usb_start(). > > > > Reported-by: syzbot+57281c762a3922e14dfe@syzkaller.appspotmail.com > > Signed-off-by: Bui Quang Minh <minhquangbui99@gmail.com> > > --- > > v1: add memory leak fix when not resubmitting urb > > v2: add memory leak fix when failing to resubmit urb > > > > drivers/net/can/usb/mcba_usb.c | 11 ++++++++--- > > 1 file changed, 8 insertions(+), 3 deletions(-) > > > > diff --git a/drivers/net/can/usb/mcba_usb.c b/drivers/net/can/usb/mcba_usb.c > > index df54eb7d4b36..30236e640116 100644 > > --- a/drivers/net/can/usb/mcba_usb.c > > +++ b/drivers/net/can/usb/mcba_usb.c > > @@ -584,6 +584,8 @@ static void mcba_usb_read_bulk_callback(struct urb *urb) > > case -EPIPE: > > case -EPROTO: > > case -ESHUTDOWN: > > + usb_free_coherent(urb->dev, urb->transfer_buffer_length, > > + urb->transfer_buffer, urb->transfer_dma); > > return; > > > > Can you call usb_free_coherent() in what can be hard IRQ context? You are right, I digged in the code and saw some comments that on some architectures, usb_free_coherent() cannot be called in hard IRQ context. I see the usb_free_coherent() is called in write_bulk_callback too. I will send a patch that uses usb_anchor to keep track of these urbs and cleanup the transfer buffer later in disconnect(). Thank you for your review, Quang Minh. ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb 2021-01-11 14:31 ` Bui Quang Minh @ 2021-01-12 6:42 ` Minh Bùi Quang 2021-01-21 15:19 ` Bui Quang Minh 0 siblings, 1 reply; 5+ messages in thread From: Minh Bùi Quang @ 2021-01-12 6:42 UTC (permalink / raw) To: Oliver Neukum Cc: linux-usb, a.darwish, bigeasy, Greg Kroah-Hartman, linux-kernel, Alan Stern, syzkaller-bugs, Thomas Gleixner On Mon, Jan 11, 2021 at 9:31 PM Bui Quang Minh <minhquangbui99@gmail.com> wrote: > > On Mon, Jan 11, 2021 at 01:00:31PM +0100, Oliver Neukum wrote: > > Am Montag, den 11.01.2021, 10:49 +0000 schrieb Bui Quang Minh: > > > In mcba_usb_read_bulk_callback(), when we don't resubmit or fails to > > > resubmit the urb, we need to deallocate the transfer buffer that is > > > allocated in mcba_usb_start(). > > > > > > Reported-by: syzbot+57281c762a3922e14dfe@syzkaller.appspotmail.com > > > Signed-off-by: Bui Quang Minh <minhquangbui99@gmail.com> > > > --- > > > v1: add memory leak fix when not resubmitting urb > > > v2: add memory leak fix when failing to resubmit urb > > > > > > drivers/net/can/usb/mcba_usb.c | 11 ++++++++--- > > > 1 file changed, 8 insertions(+), 3 deletions(-) > > > > > > diff --git a/drivers/net/can/usb/mcba_usb.c b/drivers/net/can/usb/mcba_usb.c > > > index df54eb7d4b36..30236e640116 100644 > > > --- a/drivers/net/can/usb/mcba_usb.c > > > +++ b/drivers/net/can/usb/mcba_usb.c > > > @@ -584,6 +584,8 @@ static void mcba_usb_read_bulk_callback(struct urb *urb) > > > case -EPIPE: > > > case -EPROTO: > > > case -ESHUTDOWN: > > > + usb_free_coherent(urb->dev, urb->transfer_buffer_length, > > > + urb->transfer_buffer, urb->transfer_dma); > > > return; > > > > > > > Can you call usb_free_coherent() in what can be hard IRQ context? > > You are right, I digged in the code and saw some comments that on some > architectures, usb_free_coherent() cannot be called in hard IRQ context. > I see the usb_free_coherent() is called in write_bulk_callback too. I will > send a patch that uses usb_anchor to keep track of these urbs and cleanup > the transfer buffer later in disconnect(). Hi, I have sent a version 3 patch. However, I found out that usb_free_coherent() is ok in this situation. In usb_free_coherent(), if the buffer is allocated via dma_alloc_coherent() in usb_alloc_coherent(), dma_free_coherent() is called. In dma_free_coherent(), ops->free() may be called in some cases which may contains calls to vunmap() that is not permitted in interrupt context. However, in usb_alloc_coherent(), buffer can be allocated from dma pool if the size is less than 2048 and the buffer size in mcba_usb is obviously less than 2048. As a result, usb_free_coherent() will at most fall in the path that calls dma_pool_free(), which is safe. Am I right? Thanks, Quang Minh. ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb 2021-01-12 6:42 ` Minh Bùi Quang @ 2021-01-21 15:19 ` Bui Quang Minh 0 siblings, 0 replies; 5+ messages in thread From: Bui Quang Minh @ 2021-01-21 15:19 UTC (permalink / raw) To: linux-usb Cc: a.darwish, bigeasy, Greg Kroah-Hartman, linux-kernel, Alan Stern, syzkaller-bugs, Thomas Gleixner, Oliver Neukum, Wolfgang Grandegger, Marc Kleine-Budde On Tue, Jan 12, 2021 at 01:42:33PM +0700, Minh Bùi Quang wrote: > On Mon, Jan 11, 2021 at 9:31 PM Bui Quang Minh <minhquangbui99@gmail.com> wrote: > > > > On Mon, Jan 11, 2021 at 01:00:31PM +0100, Oliver Neukum wrote: > > > Am Montag, den 11.01.2021, 10:49 +0000 schrieb Bui Quang Minh: > > > > In mcba_usb_read_bulk_callback(), when we don't resubmit or fails to > > > > resubmit the urb, we need to deallocate the transfer buffer that is > > > > allocated in mcba_usb_start(). > > > > > > > > Reported-by: syzbot+57281c762a3922e14dfe@syzkaller.appspotmail.com > > > > Signed-off-by: Bui Quang Minh <minhquangbui99@gmail.com> > > > > --- > > > > v1: add memory leak fix when not resubmitting urb > > > > v2: add memory leak fix when failing to resubmit urb > > > > > > > > drivers/net/can/usb/mcba_usb.c | 11 ++++++++--- > > > > 1 file changed, 8 insertions(+), 3 deletions(-) > > > > > > > > diff --git a/drivers/net/can/usb/mcba_usb.c b/drivers/net/can/usb/mcba_usb.c > > > > index df54eb7d4b36..30236e640116 100644 > > > > --- a/drivers/net/can/usb/mcba_usb.c > > > > +++ b/drivers/net/can/usb/mcba_usb.c > > > > @@ -584,6 +584,8 @@ static void mcba_usb_read_bulk_callback(struct urb *urb) > > > > case -EPIPE: > > > > case -EPROTO: > > > > case -ESHUTDOWN: > > > > + usb_free_coherent(urb->dev, urb->transfer_buffer_length, > > > > + urb->transfer_buffer, urb->transfer_dma); > > > > return; > > > > > > > > > > Can you call usb_free_coherent() in what can be hard IRQ context? > > > > You are right, I digged in the code and saw some comments that on some > > architectures, usb_free_coherent() cannot be called in hard IRQ context. > > I see the usb_free_coherent() is called in write_bulk_callback too. I will > > send a patch that uses usb_anchor to keep track of these urbs and cleanup > > the transfer buffer later in disconnect(). > > Hi, I have sent a version 3 patch. However, I found out that usb_free_coherent() > is ok in this situation. In usb_free_coherent(), if the buffer is allocated via > dma_alloc_coherent() in usb_alloc_coherent(), dma_free_coherent() is called. > In dma_free_coherent(), ops->free() may be called in some cases which may > contains calls to vunmap() that is not permitted in interrupt context. However, > in usb_alloc_coherent(), buffer can be allocated from dma pool if the > size is less > than 2048 and the buffer size in mcba_usb is obviously less than 2048. > As a result, > usb_free_coherent() will at most fall in the path that calls > dma_pool_free(), which is > safe. Am I right? Hi, I'm CC'ing CAN network driver maintainers so we can discuss the patch properly. I'm so sorry for spamming emails. Thanks, Quang Minh. ^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2021-01-21 15:22 UTC | newest] Thread overview: 5+ messages (download: mbox.gz follow: Atom feed -- links below jump to the message on this page -- 2021-01-11 10:49 [PATCH v2] can: mcba_usb: Fix memory leak when cancelling urb Bui Quang Minh 2021-01-11 12:00 ` Oliver Neukum 2021-01-11 14:31 ` Bui Quang Minh 2021-01-12 6:42 ` Minh Bùi Quang 2021-01-21 15:19 ` Bui Quang Minh
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox; as well as URLs for NNTP newsgroup(s).