Re: [PATCH] dmaengine/dmatest: terminate transfers only in case of errors

[Jon Mason <jon.mason@intel.com>]

On Fri, Nov 09, 2012 at 06:44:18PM +0000, Dan Williams wrote:
> 
> 
> On 11/9/12 7:26 AM, "Viresh Kumar" <viresh.kumar@linaro.org> wrote:
> 
> >From: Shiraz Hashim <shiraz.hashim@st.com>
> >
> >dmatest erroneously terminated transfers in normal cases also leading to
> >test failures for multiple threads over a channel. Fix this and
> >terminate transfers only in case of errors.
> >
> >Signed-off-by: Shiraz Hashim <shiraz.hashim@st.com>
> >Signed-off-by: Deepak Sikri <deepak.sikri@st.com>
> >---
> > drivers/dma/dmatest.c | 4 +++-
> > 1 file changed, 3 insertions(+), 1 deletion(-)
> >
> >diff --git a/drivers/dma/dmatest.c b/drivers/dma/dmatest.c
> >index 8f130d4e..ce58b8e 100644
> >--- a/drivers/dma/dmatest.c
> >+++ b/drivers/dma/dmatest.c
> >@@ -509,7 +509,9 @@ err_srcs:
> > 			thread_name, total_tests, failed_tests, ret);
> > 
> > 	/* terminate all transfers on specified channels */
> >-	chan->device->device_control(chan, DMA_TERMINATE_ALL, 0);
> >+	if (ret)
> >+		chan->device->device_control(chan, DMA_TERMINATE_ALL, 0);
> >+
> 
> This and commit 9704efa "dmaengine/dmatest: Terminate transfers on all
> channels in case of error or exit" are broken for channels that don't
> implement the optional ->device_control.  Care to fix that up?

I think the best way is to call dmaengine_terminate_all() and have a
sanity check in dmaengine_device_control() that checks for a NULL
device_control.  I have a patch that does this and fixes the NULL
pointer error I am seeing on IOAT.  I will send it out shortly.

Thanks,
Jon

> 
> 
> --
> Dan
>