On 02/17/2016 06:24 PM, Bart Van Assche wrote:
> On 02/17/2016 07:57 AM, Doug Ledford wrote:
>> On 02/11/2016 02:08 PM, Bart Van Assche wrote:
>>> To make it possible to determine why an initiator sent a REJ,
>>> log the private data associated with the received REJ packet.
>>
>> This particular function is all sorts of broken...I fixed it up as I
>> took it in. But, issues were:
>>
>> 1) We always allocate priv even when private_data_len is 0
>> 2) Since we always allocate priv, the test for priv in the pr_info args
>> isn't really helpful, but the way the test is set up, if priv exists,
>> then we actually pass NULL, if it doesn't, we pass the static string.
>> 3) In the event we have priv and private data, then there is a nit about
>> formatting because priv will end with a space right before the null
>> terminator, which makes the period in the pr_info format hang out in
>> open space.
>>
>> My changes:
>>
>> 1) Only alloc priv if private_data_len is not 0
>> 2) Change priv to be space prepended instead of space postpended
>> 3) Change pr_info format to work with space prepended string
>> 4) Fix test in pr_info args
>
> Hello Doug,
>
> Thanks for stepping in and addressing this. But as far as I can see
> these changes have not yet been pushed out to the k.o/for-4.6 branch on
> git://git.kernel.org/pub/scm/linux/kernel/git/dledford/rdma.git ? I will
> have another look once your tree is publicly available.
>
> Bart.
It's in my srpt branch at the moment. I haven't pulled the topic
branches into my k.o/for-4.6 branch yet (when I pushed it last night to
github it was to get 0day testing before committing it to my k.o branch
and pushing it to k.o).
--
Doug Ledford <dledford@xxxxxxxxxx>
GPG KeyID: 0E572FDD
Attachment:
signature.asc
Description: OpenPGP digital signature
