| Message ID | 20170725205753.4735-6-logang@deltatee.com (mailing list archive) |
|---|---|
| State | New, archived |
| Delegated to: | Bjorn Helgaas |
| Headers | show |
On Tue, Jul 25, 2017 at 02:57:42PM -0600, Logan Gunthorpe wrote: > After the link tests, there is a race on one side of the test for > the link coming up. It's possible, in some cases, for the test script > to write to the 'peer_trans' files before the link has come up. > > To fix this, we simply use the link event file to ensure both sides > see the link as up before continuning. > > Signed-off-by: Logan Gunthorpe <logang@deltatee.com> This looks like a bug fix. Assuming this is the case, I can pull it out, add a "Fixes" line, and add it to my bug fixes branch. Is this the case? Thanks, Jon > --- > tools/testing/selftests/ntb/ntb_test.sh | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/tools/testing/selftests/ntb/ntb_test.sh b/tools/testing/selftests/ntb/ntb_test.sh > index 1c12b5855e4f..5fc7ad359e21 100755 > --- a/tools/testing/selftests/ntb/ntb_test.sh > +++ b/tools/testing/selftests/ntb/ntb_test.sh > @@ -333,6 +333,10 @@ function ntb_tool_tests() > link_test $LOCAL_TOOL $REMOTE_TOOL > link_test $REMOTE_TOOL $LOCAL_TOOL > > + #Ensure the link is up on both sides before continuing > + write_file Y $LOCAL_TOOL/link_event > + write_file Y $REMOTE_TOOL/link_event > + > for PEER_TRANS in $(ls $LOCAL_TOOL/peer_trans*); do > PT=$(basename $PEER_TRANS) > write_file $MW_SIZE $LOCAL_TOOL/$PT > -- > 2.11.0 > > -- > You received this message because you are subscribed to the Google Groups "linux-ntb" group. > To unsubscribe from this group and stop receiving emails from it, send an email to linux-ntb+unsubscribe@googlegroups.com. > To post to this group, send email to linux-ntb@googlegroups.com. > To view this discussion on the web visit https://groups.google.com/d/msgid/linux-ntb/20170725205753.4735-6-logang%40deltatee.com. > For more options, visit https://groups.google.com/d/optout.
On 01/08/17 12:07 PM, Jon Mason wrote: > On Tue, Jul 25, 2017 at 02:57:42PM -0600, Logan Gunthorpe wrote: >> After the link tests, there is a race on one side of the test for >> the link coming up. It's possible, in some cases, for the test script >> to write to the 'peer_trans' files before the link has come up. >> >> To fix this, we simply use the link event file to ensure both sides >> see the link as up before continuning. >> >> Signed-off-by: Logan Gunthorpe <logang@deltatee.com> > > This looks like a bug fix. Assuming this is the case, I can pull it > out, add a "Fixes" line, and add it to my bug fixes branch. Is this > the case? Sure, yup, if you'd like to do that I'm fine with it. Technically, I don't think the bug can be triggered until the patches later in the series are applied. Thanks, Logan
On Tue, Aug 01, 2017 at 12:09:18PM -0600, Logan Gunthorpe wrote: > > On 01/08/17 12:07 PM, Jon Mason wrote: > > On Tue, Jul 25, 2017 at 02:57:42PM -0600, Logan Gunthorpe wrote: > >> After the link tests, there is a race on one side of the test for > >> the link coming up. It's possible, in some cases, for the test script > >> to write to the 'peer_trans' files before the link has come up. > >> > >> To fix this, we simply use the link event file to ensure both sides > >> see the link as up before continuning. > >> > >> Signed-off-by: Logan Gunthorpe <logang@deltatee.com> > > > > This looks like a bug fix. Assuming this is the case, I can pull it > > out, add a "Fixes" line, and add it to my bug fixes branch. Is this > > the case? > > Sure, yup, if you'd like to do that I'm fine with it. Technically, I > don't think the bug can be triggered until the patches later in the > series are applied. Given how trivial it is, I think closing the loop here on this would be a good thing (and one less patch for your v4). > > Thanks, > > Logan > > -- > You received this message because you are subscribed to the Google Groups "linux-ntb" group. > To unsubscribe from this group and stop receiving emails from it, send an email to linux-ntb+unsubscribe@googlegroups.com. > To post to this group, send email to linux-ntb@googlegroups.com. > To view this discussion on the web visit https://groups.google.com/d/msgid/linux-ntb/46065aa6-0fe8-59d5-63dc-3beb66b69154%40deltatee.com. > For more options, visit https://groups.google.com/d/optout.
On 01/08/17 12:16 PM, Jon Mason wrote: > On Tue, Aug 01, 2017 at 12:09:18PM -0600, Logan Gunthorpe wrote: >> >> On 01/08/17 12:07 PM, Jon Mason wrote: >>> On Tue, Jul 25, 2017 at 02:57:42PM -0600, Logan Gunthorpe wrote: >>>> After the link tests, there is a race on one side of the test for >>>> the link coming up. It's possible, in some cases, for the test script >>>> to write to the 'peer_trans' files before the link has come up. >>>> >>>> To fix this, we simply use the link event file to ensure both sides >>>> see the link as up before continuning. >>>> >>>> Signed-off-by: Logan Gunthorpe <logang@deltatee.com> >>> >>> This looks like a bug fix. Assuming this is the case, I can pull it >>> out, add a "Fixes" line, and add it to my bug fixes branch. Is this >>> the case? >> >> Sure, yup, if you'd like to do that I'm fine with it. Technically, I >> don't think the bug can be triggered until the patches later in the >> series are applied. > > Given how trivial it is, I think closing the loop here on this would > be a good thing (and one less patch for your v4). Sounds good to me. I'll rebase v4 onto ntb_next again once I see the patch in. I also have yet to rename the file per Dave's feedback. Once I've done both those things I'll send a v4. Thanks, Logan
diff --git a/tools/testing/selftests/ntb/ntb_test.sh b/tools/testing/selftests/ntb/ntb_test.sh index 1c12b5855e4f..5fc7ad359e21 100755 --- a/tools/testing/selftests/ntb/ntb_test.sh +++ b/tools/testing/selftests/ntb/ntb_test.sh @@ -333,6 +333,10 @@ function ntb_tool_tests() link_test $LOCAL_TOOL $REMOTE_TOOL link_test $REMOTE_TOOL $LOCAL_TOOL + #Ensure the link is up on both sides before continuing + write_file Y $LOCAL_TOOL/link_event + write_file Y $REMOTE_TOOL/link_event + for PEER_TRANS in $(ls $LOCAL_TOOL/peer_trans*); do PT=$(basename $PEER_TRANS) write_file $MW_SIZE $LOCAL_TOOL/$PT
After the link tests, there is a race on one side of the test for the link coming up. It's possible, in some cases, for the test script to write to the 'peer_trans' files before the link has come up. To fix this, we simply use the link event file to ensure both sides see the link as up before continuning. Signed-off-by: Logan Gunthorpe <logang@deltatee.com> --- tools/testing/selftests/ntb/ntb_test.sh | 4 ++++ 1 file changed, 4 insertions(+)