April 2013 - Darwin-dev - lists.apple.com

Re: About new os x 10.8.3 source code
by Todd Heberlein 12 Apr '13

12 Apr '13

On Apr 12, 2013, at 7:19 AM, William Siegrist <wsiegrist(a)apple.com> wrote: > Sorry for the delay. We are working on the release and hope to have it out soon. Just had a "Doh!" moment. Does it help when filing bug reports to review the source code to see if we can pin down the underlying cause of the problem (and optionally test a fix)? That is, is the bug more likely to get fixed a little faster doing this? I filed a bug report that probably affects very few people (the user's launchd process isn't generating BSM audit records properly; I suspect it isn't setting its audit mask correctly), and I hadn't thought of looking through the source code before filing a bug report. Should I do this in the future? Thanks, Todd _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.ap… This email sent to site_archiver(a)lists.apple.com

1 0

Re: About new os x 10.8.3 source code
by William Siegrist 12 Apr '13

12 Apr '13

_______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.ap… This email sent to site_archiver(a)lists.apple.com

1 0

About new os x 10.8.3 source code
by Sopl'Wang 12 Apr '13

12 Apr '13

_______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.ap… This email sent to site_archiver(a)lists.apple.com

1 0

Anyone know how to disable VT-d ?
by Mark Gilbert 05 Apr '13

05 Apr '13

Folks. We have a problem with MacMini 2012 Quad i7, which I believe may be related to Intel VT-d virtualisation on this machine. The previous generation Quad i7 mini server does not suffer the problem and does not support VT-d The question is - is there any fairly easy way to disable VT-d on the new machine ? Thanks _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.ap… This email sent to site_archiver(a)lists.apple.com

1 0

Re: Which version of SMB is in 10.8.2 ? Looks like a contradiction
by Shantonu Sen 03 Apr '13

03 Apr '13

_______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.ap… This email sent to site_archiver(a)lists.apple.com

1 0

Which version of SMB is in 10.8.2 ? Looks like a contradiction
by Mark Gilbert 03 Apr '13

03 Apr '13

Folks. I am attempting to solve a very deep file system problem we are experiencing, and in doing so I am digging into the OSX source code published at opensource.apple.com - The machine I am testing with is a late 2012 MacMini with 10.8.2, and during the failure, we see the error: kernel[0]: smb_iod_sendrq: TRAN_SEND returned non-fatal error 35 sr_cmd = ..(a whole variety) Looking at opensource.apple.com suggests that 10.8.2 includes smb-583.8, but examining the function smb_iod_sendrq from the 583.8 source code reveals that the error message is just : smb_iod_sendrq: TRAN_SEND returned non-fatal error %d ie. it doesn't suffix the "sr_cmd=" string However, if you look at older versions of SMB, you will find the exact error message (including the sr_cmd string, included in smb-551 and some older versions. >From this I can only conclude that the 10.8.2 OS which we are testing with does NOT use smb-583.8 as the opensource.apple website suggests. Could anyone comment on this ????? For anyone interested, we are seeing a 100% repeatable situation where our application can essentially stall the entire SMB client environment to the point of failure, even after our app is quit. All our app does is read and write some files (at the same time). Fails 100%: - MacMini late 2012, SMB connection to another MacMini Server Fails 100%: - MacMini late 2012, NFS connection to Isilon Storage Fails NEVER - MacMini late 2012, AFP connection to another MacMini Server Fails NEVER - MacMini mid 2011 any storage, any protocol Interestingly the Isilon NFS failure also shows error 35 (EAGAIN - resource unavailable) so this problem is not specific to SMB. Interesting that AFP never, ever fails, and mid 2011 Mac Mini with same OS. never fails. Been going around in circles for 2 weeks on this, tried everything. It just seems our sequence of file calls takes down SMB or NFS, but ONLY with these new late 2012 Mac Minis..... wierd Any ideas ? Thanks _______________________________________________ Do not post admin requests to the list. They will be ignored. Darwin-dev mailing list (Darwin-dev(a)lists.apple.com) Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/darwin-dev/site_archiver%40lists.ap… This email sent to site_archiver(a)lists.apple.com

1 0