diff options
author | 2009-06-10 14:44:14 -0600 | |
---|---|---|
committer | 2009-07-12 13:21:44 -0700 | |
commit | aea0d43bdec32cc4a7ba53d6c82616de3964357a (patch) | |
tree | a242512399606fe497c2f6977e17d4ffccec6945 /lib/fault-inject.c | |
parent | Staging: vt6655: compile fix (diff) | |
download | wireguard-linux-aea0d43bdec32cc4a7ba53d6c82616de3964357a.tar.xz wireguard-linux-aea0d43bdec32cc4a7ba53d6c82616de3964357a.zip |
Staging: rspiusb: use NULL virtual address instead of a bogus one
The main problem here is that I just cannot see how this could ever
be correct:
usb_fill_bulk_urb(pdx->PixelUrb[frameInfo][i], pdx->udev, epAddr,
(dma_addr_t *) sg_dma_address(&pdx->sgl[frameInfo][i]),
sg_dma_len(&pdx->sgl[frameInfo][i]),
You cannot take a DMA address, cast it to a _pointer to_ a DMA address,
and then regard it as a virtual address of the transfer buffer.
However, finding the right virtual address was too hard for me,
so I just stubbed it with NULL. At least usbmon won't oops then
(it will not show any data but it's better than crashing).
Also, too big a buffer was allocated elsewhere.
And since we're at it, drop clearly unnecessary usb_buffer_alloc too,
leaving it where it may be useful.
Signed-off-by: Pete Zaitcev <zaitcev@redhat.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de>
Diffstat (limited to 'lib/fault-inject.c')
0 files changed, 0 insertions, 0 deletions