Re: [Qemu-devel] [PATCH] intel-hda: fix stream search

[Kevin Wolf <kwolf@redhat.com>]

Am 02.11.2011 13:18, schrieb Andreas Färber:
> Am 02.11.2011 12:56, schrieb Gerd Hoffmann:
>> commit ba43d28916c4f51c19bd7366089155ce81bee058 introduces a bug:
>> The stream-not-found case doesn't error out any more, instead the
>> code silently uses the first stream.  Fix it.
>>
>> Signed-off-by: Gerd Hoffmann <kraxel@redhat.com>
>> ---
>>  hw/intel-hda.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/hw/intel-hda.c b/hw/intel-hda.c
>> index 675b659..b7ec36d 100644
>> --- a/hw/intel-hda.c
>> +++ b/hw/intel-hda.c
>> @@ -401,7 +401,7 @@ static bool intel_hda_xfer(HDACodecDevice *dev, uint32_t stnr, bool output,
>>              break;
>>          }
>>      }
>> -    if (st == NULL) {
>> +    if (s == 4) {
>>          return false;
>>      }
>>      if (st->bpl == NULL) {
> 
> The commit message doesn't say it clearly: Is st == NULL impossible?
> If not, this introduces a potential NULL pointer dereference at st->bpl.

Yes, it's impossible. This is the full context:

    st = output ? d->st + 4 : d->st;
    for (s = 0; s < 4; s++) {
        if (stnr == ((st[s].ctl >> 20) & 0x0f)) {
            st = st + s;
            break;
        }
    }

d->st is an array: IntelHDAStream st[8];

Kevin