Message ID | 20190628072357.31782-1-kraxel@redhat.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | virtio-gpu: fix unmap in error path | expand |
On 06/28/19 09:23, Gerd Hoffmann wrote: > We land here in case not everything we've asked for could be mapped. > So unmap only the bytes which have actually been mapped. > > Also we didn't access anything, so acces_len can be 0. s/acces_len/access_len/ With that: Reviewed-by: Laszlo Ersek <lersek@redhat.com> Thank you! Laszlo > > Reported-by: Laszlo Ersek <lersek@redhat.com> > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com> > --- > hw/display/virtio-gpu.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/hw/display/virtio-gpu.c b/hw/display/virtio-gpu.c > index 2b0f66b1d68d..475a018c027c 100644 > --- a/hw/display/virtio-gpu.c > +++ b/hw/display/virtio-gpu.c > @@ -1048,9 +1048,9 @@ static int virtio_gpu_load(QEMUFile *f, void *opaque, size_t size, > if (res->iov[i].iov_base) { > dma_memory_unmap(VIRTIO_DEVICE(g)->dma_as, > res->iov[i].iov_base, > - res->iov[i].iov_len, > + len, > DMA_DIRECTION_TO_DEVICE, > - res->iov[i].iov_len); > + 0); > } > /* ...and the mappings for previous loop iterations */ > res->iov_cnt = i; >
Gerd Hoffmann <kraxel@redhat.com> 于2019年6月28日周五 下午3:24写道: > We land here in case not everything we've asked for could be mapped. > So unmap only the bytes which have actually been mapped. > > Also we didn't access anything, so acces_len can be 0. > > Reported-by: Laszlo Ersek <lersek@redhat.com> > Signed-off-by: Gerd Hoffmann <kraxel@redhat.com> > It is reasonable, so Reviewed-by: Li Qiang <liq3ea@gmail.com> btw: Does it break something before this patch? AFAICS, the 'len' is not used for the unmap, only the 'access_len' is used. Thanks, Li Qiang > --- > hw/display/virtio-gpu.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/hw/display/virtio-gpu.c b/hw/display/virtio-gpu.c > index 2b0f66b1d68d..475a018c027c 100644 > --- a/hw/display/virtio-gpu.c > +++ b/hw/display/virtio-gpu.c > @@ -1048,9 +1048,9 @@ static int virtio_gpu_load(QEMUFile *f, void > *opaque, size_t size, > if (res->iov[i].iov_base) { > dma_memory_unmap(VIRTIO_DEVICE(g)->dma_as, > res->iov[i].iov_base, > - res->iov[i].iov_len, > + len, > DMA_DIRECTION_TO_DEVICE, > - res->iov[i].iov_len); > + 0); > } > /* ...and the mappings for previous loop iterations */ > res->iov_cnt = i; > -- > 2.18.1 > > >
diff --git a/hw/display/virtio-gpu.c b/hw/display/virtio-gpu.c index 2b0f66b1d68d..475a018c027c 100644 --- a/hw/display/virtio-gpu.c +++ b/hw/display/virtio-gpu.c @@ -1048,9 +1048,9 @@ static int virtio_gpu_load(QEMUFile *f, void *opaque, size_t size, if (res->iov[i].iov_base) { dma_memory_unmap(VIRTIO_DEVICE(g)->dma_as, res->iov[i].iov_base, - res->iov[i].iov_len, + len, DMA_DIRECTION_TO_DEVICE, - res->iov[i].iov_len); + 0); } /* ...and the mappings for previous loop iterations */ res->iov_cnt = i;
We land here in case not everything we've asked for could be mapped. So unmap only the bytes which have actually been mapped. Also we didn't access anything, so acces_len can be 0. Reported-by: Laszlo Ersek <lersek@redhat.com> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com> --- hw/display/virtio-gpu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)