[ofa-general] [PATCH linux-next 2/5] RDMA/cxgb3: Don't free the endpoint early.

[Steve Wise]

Roland Dreier wrote:
>  > - Endpoint flags now need to be set via atomic bitops because they can
>  > be set on both the iw_cxgb3 workqueue thread and user disconnect threads.
>
>  > +	if (!test_bit(ABORT_REQ_IN_PROGRESS, &ep->com.flags)) {
>  > +		set_bit(ABORT_REQ_IN_PROGRESS, &ep->com.flags);
>
> for atomicity, should all the places that do test_bit then set_bit
> really be using test_and_set_bit()?
>
> it would be cleaner anyway.
>
>  - R.
>   


This particular bit is only set/read on the workq thread.  But I agree I 
should be using test_and_set_bit().   

I'll resend.


Steve.