Re: [Qemu-devel] [PATCH v2 14/17] net/dp8393x: correctly reset in_use field

[Hervé Poussineau]

Le 02/06/2015 13:04, Aurelien Jarno a écrit :
> On 2015-05-27 14:19, Hervé Poussineau wrote:
> > Don't write more than the field width, which is always 16 bit.
> > Fixes network in NetBSD 5.1/arc
> >
> > Signed-off-by: Hervé Poussineau <address@hidden>
> > ---
> >   hw/net/dp8393x.c | 4 ++--
> >   1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/hw/net/dp8393x.c b/hw/net/dp8393x.c
> > index 4184045..b72b0b1 100644
> > --- a/hw/net/dp8393x.c
> > +++ b/hw/net/dp8393x.c
> > @@ -761,10 +761,10 @@ static ssize_t dp8393x_receive(NetClientState *nc, const 
> > uint8_t * buf,
> >           /* EOL detected */
> >           s->regs[SONIC_ISR] |= SONIC_ISR_RDE;
> >       } else {
> > -        data[0 * width] = 0; /* in_use */
> > +        uint16_t in_use = 0;
> >           address_space_rw(&s->as,
> >               ((s->regs[SONIC_URDA] << 16) | s->regs[SONIC_CRDA]) + 
> > sizeof(uint16_t) * 6 * width,
> > -            MEMTXATTRS_UNSPECIFIED, (uint8_t *)data, size, 1);
> > +            MEMTXATTRS_UNSPECIFIED, (uint8_t *)&in_use, sizeof(uint16_t), 1);
>
> Why not initialising both data[0] and data[1] to 0 and a fixed size of 2
> bytes instead of using a new variable?

I'll do that if a v3 is required.

> >           s->regs[SONIC_CRDA] = s->regs[SONIC_LLFA];
> >           s->regs[SONIC_ISR] |= SONIC_ISR_PKTRX;
> >           s->regs[SONIC_RSC] = (s->regs[SONIC_RSC] & 0xff00) | (((s->regs[SONIC_RSC] 
> > & 0x00ff) + 1) & 0x00ff);
>
> That said:
> Reviewed-by: Aurelien Jarno <address@hidden>

Hervé