-----"Leon Romanovsky" <leon@xxxxxxxxxx> wrote: -----
>To: "Bernard Metzler" <bmt@xxxxxxxxxxxxxx>
>From: "Leon Romanovsky" <leon@xxxxxxxxxx>
>Date: 10/02/2019 03:32PM
>Cc: linux-rdma@xxxxxxxxxxxxxxx, bharat@xxxxxxxxxxx, jgg@xxxxxxxx,
>nirranjan@xxxxxxxxxxx, krishna2@xxxxxxxxxxx, bvanassche@xxxxxxx
>Subject: [EXTERNAL] Re: [[PATCH v2 for-next]] RDMA/siw: Fix SQ/RQ
>drain logic
>
>On Wed, Oct 02, 2019 at 03:14:23PM +0200, Bernard Metzler wrote:
>> Storage ULPs (e.g. iSER & NVMeOF) use ib_drain_qp() to
>> drain QP/CQ. Current SIW's own drain routines do not properly
>> wait until all SQ/RQ elements are completed and reaped
>> from the CQ. This may cause touch after free issues.
>> New logic relies on generic __ib_drain_sq()/__ib_drain_rq()
>> posting a final work request, which SIW immediately flushes
>> to CQ.
>>
>> Fixes: 303ae1cdfdf7 ("rdma/siw: application interface")
>> Signed-off-by: Krishnamraju Eraparaju <krishna2@xxxxxxxxxxx>
>> Signed-off-by: Bernard Metzler <bmt@xxxxxxxxxxxxxx>
>> ---
>> drivers/infiniband/sw/siw/siw_main.c | 20 -----
>> drivers/infiniband/sw/siw/siw_verbs.c | 103
>+++++++++++++++++++++-----
>> 2 files changed, 86 insertions(+), 37 deletions(-)
>>
>
>I didn't follow after v1 discussion and will be glad to see
>changelog,
>what is the reason for v2?
>
Absolutely. Let me resend the patch with below change log.
Sorry & best regards,
Bernard.
v1 -> v2:
- Accept SQ and RQ work requests, if QP is in ERROR
state. In that case, immediately flush WR's to CQ.
This already provides needed functionality to
support ib_drain_sq()/ib_drain_rq() without extra
state checking in the fast path.
