[ofw] [PATCH 1/3] IBAL (checksum): restoring vendor_specific field
Smith, Stan
stan.smith at intel.com
Tue Sep 9 08:53:42 PDT 2008
Alex Naslednikov wrote:
> Hello Fab,
> I received you proposal after the patch was implemented. During the
> testing, we found a lot of additional issues and we were motivated to
> finish the testing it time in order to allow our changes to be placed
> in
> 2.0.0 branch.
> Each additional change (even the smallest one) requires us to check
> again all the flow on : 2-3 types of Windows x 2 types of architecture
> (x64, x86) x 3 types of HCAs (at least) x 3 types of option values.
> Of course, your proposal to use receive_opt field make sense and we
> are totally agree with it.
> So we will insert it into trunk as soon as possible and will carefully
> check it on all the platforms.
Hello,
Understanding the amount of testing required, it would be good to include this in WinOF 2.0.
Do you foresee that happening?
If so, perhaps this week or next week (RC2)?
Thanks,
Stan.
>
> Thanks,
> XaleX
>
> -----Original Message-----
> From: Fab Tillier [mailto:ftillier at windows.microsoft.com]
> Sent: Monday, September 08, 2008 7:14 PM
> To: Tzachi Dar; Alex Naslednikov; ofw at lists.openfabrics.org
> Subject: RE: [ofw] [PATCH 1/3] IBAL (checksum): restoring
> vendor_specific field
>
>> Applied to trunk and branch.
>
> I guess this means that the checksum values aren't going to be encoded
> in the existing receive options field?
>
> It would be really nice to get some feedback on comments made to the
> list. Just posting patches, waiting a week or so without responding
> to questions, and then checking in doesn't foster a sense of
> community and cooperation.
>
> -Fab
>
>>
>> Thanks
>> Tzachi
>>
>>> -----Original Message-----
>>> From: Alex Naslednikov
>>> Sent: Thursday, September 04, 2008 7:34 PM
>>> To: Tzachi Dar; ofw at lists.openfabrics.org
>>> Cc: Ishai Rabinovitz
>>> Subject: [ofw] [PATCH 1/3] IBAL (checksum): restoring
>>> vendor_specific
>
>>> field
>>>
>>> Signed-off by: xalex (Alexander Naslednikov)
>>> Index:
>>> D:/Windows/MLNX_WINOF_CSUM_ARBEL/ulp/opensm/user/include/iba/i
>>> b_types_extended.h
>>> ===================================================================
>>> --- D:/Windows/MLNX_WINOF_CSUM_ARBEL/ulp/opensm/user/include/iba/i
>>> b_types_extended.h (revision 3104) +++
>>> D:/Windows/MLNX_WINOF_CSUM_ARBEL/ulp/opensm/user/include/iba/i
>>> b_types_extended.h (revision 3105)
>>> @@ -2096,12 +2096,7 @@
>>> ib_wc_type_t wc_type;
>>>
>>> uint32_t length;
>>> - struct {
>>> - uint8_t vendor_specific;
>>> - uint8_t csum_ok;
>>> - uint16_t vendor_specific2;
>>> - uint32_t vendor_specific3;
>>> - };
>>> + uint64_t vendor_specific;
>>> ib_wc_status_t status;
>>>
>>> union _wc_recv
>>> @@ -2122,6 +2117,7 @@
>>> ib_net16_t remote_lid;
>>> uint8_t remote_sl;
>>> uint8_t path_bits;
>>> + uint8_t csum_ok;
>>>
>>> } ud;
>>>
>>> Index: D:/Windows/MLNX_WINOF_CSUM_ARBEL/inc/iba/ib_types.h
>>> ===================================================================
>>> --- D:/Windows/MLNX_WINOF_CSUM_ARBEL/inc/iba/ib_types.h (revision
>>> 3104) +++ D:/Windows/MLNX_WINOF_CSUM_ARBEL/inc/iba/ib_types.h
>>> (revision 3105) @@ -10644,12 +10644,7 @@
>>> ib_wc_type_t wc_type;
>>>
>>> uint32_t length;
>>> - struct {
>>> - uint8_t vendor_specific;
>>> - uint8_t csum_ok;
>>> - uint16_t vendor_specific2;
>>> - uint32_t vendor_specific3;
>>> - };
>>> + uint64_t vendor_specific;
>>> ib_wc_status_t status;
>>>
>>> union _wc_recv
>>> @@ -10670,6 +10665,7 @@
>>> ib_net16_t remote_lid;
>>> uint8_t remote_sl;
>>> uint8_t path_bits;
>>> + uint8_t csum_ok;
>>>
>>> } ud;
>>>
>> _______________________________________________
>> ofw mailing list
>> ofw at lists.openfabrics.org
>> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw
> _______________________________________________
> ofw mailing list
> ofw at lists.openfabrics.org
> http://lists.openfabrics.org/cgi-bin/mailman/listinfo/ofw
More information about the ofw
mailing list