12:10:15 <andyf> tsoome - I know the patch works to build gate, but I need to check that things like the DTD validation is working as it did before in zonecfg, libbrand, libpool (I probably won't test isnsd!)
12:20:31 <tsoome> aye.
15:58:00 <sommerfeld> tsoome: possibly dumb question about 15992: if I'm reading the patch correctly, it suggests that anyone using that a client using FileAllInformation would also get unexpected values in the "position", "mode", and "alignment" fields that are in between "access" and "name", offset by whatever the size of the FileAccessInformation is.   if that's the case the bug should mention that, too..
17:52:10 <tsoome_> mm, yes, apparently so, yes. because we do keep track of mbuf chain offset and increment it on each write, so missed data means the rest will be stored at wrong offset.
18:02:15 <sommerfeld> also looks like the erroneous comment with the missing field ordering was copied to the smbfs client.   Luckily the client side code doesn't read the misplaced data
18:09:30 <sommerfeld> (which I think suspect answers the question of why this wasn't noticed sooner: likely few clients actually go that deep into the message..)
18:19:03 <tsoome_> https://code.illumos.org/c/illumos-gate/+/3182 :)
18:19:04 <fenix> → CODE REVIEW 3182: 16112 smbfs: update comment about FileAllInformation (NEW) | https://www.illumos.org/issues/16112
18:20:23 <sommerfeld> Thanks!
18:23:25 <tsoome_> I guess I'll get it to RTI as is:D
20:46:51 <gitomat> [illumos-gate] 16055 truss should not be isaexec'd -- Richard Lowe <richlowe⊙rn>