On Thu, 13 Apr 2006 23:02:04 +0200 (CEST) Guennadi Liakhovetski wrote:
> On Thu, 13 Apr 2006, Randy.Dunlap wrote:
>
> > On Sun, 2 Apr 2006 21:57:43 +0200 (CEST) Guennadi Liakhovetski wrote:
> >
> > > Hello all,
> > >
> > > This time I really mean it:-)
> >
> > Hi Guennadi,
> > James pointed me at this patch since I have aha152x also having
> > problems with using page_address() while trying to do PIO.
> > I see a NULL pointer during a data-out phase. I'll attempt
> > to use the kmap_atomic() support that you have added in the
> > aha152x driver.
>
> Would be great to put these functions under some more extensive testing!
> I also want to use them in tmscsim, but there I do not have a test case
> for PIO:-(
I have a patch that seems to work for the data-in phase of the
aha152x (_cs) driver. However, the data-out phase is causing
data corruption. (I'm only using it on a test target, so no big
deal.)
Does anyone know/recall this condition in aha152x.c::datao_end() ?
("bytes to resend"):
if(TESTLO(DMASTAT, DFIFOEMP)) {
int data_count = (DATA_LEN - CURRENT_SC->resid) - GETSTCNT();
DPRINTK(debug_datao, DEBUG_LEAD "datao: %d bytes to resend (%d written, %d transferred)\n",
CMDINFO(CURRENT_SC),
data_count,
DATA_LEN-CURRENT_SC->resid,
GETSTCNT());
CURRENT_SC->resid += data_count;
and then (if use_sg), the driver mucks with the SCp scratch/work area:
if(CURRENT_SC->use_sg) {
data_count -= CURRENT_SC->SCp.ptr - SG_ADDRESS(CURRENT_SC->SCp.buffer);
while(data_count>0) {
CURRENT_SC->SCp.buffer--;
CURRENT_SC->SCp.buffers_residual++;
data_count -= CURRENT_SC->SCp.buffer->length;
}
CURRENT_SC->SCp.ptr = SG_ADDRESS(CURRENT_SC->SCp.buffer) - data_count;
CURRENT_SC->SCp.this_residual = CURRENT_SC->SCp.buffer->length + data_count;
} else {
CURRENT_SC->SCp.ptr -= data_count;
CURRENT_SC->SCp.this_residual += data_count;
}
}
Does modifying the SCp: buffer, buffers_residual, ptr, this_residual...
tell the SCSI mid-layer anything? or does this tell the driver
internally where to restart a data transfer? (if so, where is that,
if anyone knows. I can't find it.)
---
~Randy
-
: send the line "unsubscribe linux-scsi" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
