Page MenuHomeFreeBSD

dgmorris_earthlink.net (Don Morris)
User

Projects

User does not belong to any projects.

User Details

User Since
Feb 28 2018, 7:48 PM (190 w, 6 d)

Recent Activity

Apr 30 2021

dgmorris_earthlink.net requested review of D30060: Flag hn and storvsc statistics with CTLFLAG_STATS..
Apr 30 2021, 6:34 PM

Mar 3 2021

dgmorris_earthlink.net accepted D29046: uma: Fall back to other domains even for FIRSTTOUCH.
Mar 3 2021, 9:07 PM
dgmorris_earthlink.net updated the diff for D29045: Don't use sleeping allocations for ufs dirhash blocks when holding directory vnode.

Took Mark's suggestion on dropping _NOWAIT suffix on allocation macro.
Left the macro for symmetry with the FREE macros in the file.

Mar 3 2021, 8:09 PM
dgmorris_earthlink.net added inline comments to D29046: uma: Fall back to other domains even for FIRSTTOUCH.
Mar 3 2021, 8:02 PM
dgmorris_earthlink.net requested changes to D29046: uma: Fall back to other domains even for FIRSTTOUCH.
Mar 3 2021, 7:40 PM
dgmorris_earthlink.net requested review of D29045: Don't use sleeping allocations for ufs dirhash blocks when holding directory vnode.
Mar 3 2021, 6:37 PM

Dec 18 2020

dgmorris_earthlink.net accepted D27671: dtrace: Fix /"string" == NULL/ comparisons using an uninitialized value..
Dec 18 2020, 6:23 PM

Dec 15 2020

dgmorris_earthlink.net accepted D27620: acpi: Ensure that the affinity table is maximally coalesced.
Dec 15 2020, 3:53 PM

Dec 14 2020

dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

I'd have to get the machine back to go drill down, but there's some caching allocations in there -- I don't think that's out of the norm.

Dec 14 2020, 11:24 PM
dgmorris_earthlink.net accepted D27613: acpi: Ensure that entries in the affinity table are coalesced.
Dec 14 2020, 11:09 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

With D24757:

Dec 14 2020, 10:56 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

Yeah, that makes sense given what we're seeing. I'll try it out and let you know.

Dec 14 2020, 10:23 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

Since I assume the latency split may be related, I booted "-v" and pulled the SRAT lines relevant to memory (you don't need all the CPU and table stuff, I assume). Left out all the 0 addr, 0 len disabled ones as well:

Dec 14 2020, 8:36 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

Hopefully everything you need/asked for in this attachment.

Dec 14 2020, 8:15 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

No custom calls that I know of (or can find on a quick double check of the code). May end up being one I missed, of course.
The domain 1 ranges I see with the problem are avail and contiguous, just split into separate ranges -- I'll get you the output you requested.

Dec 14 2020, 5:27 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

So -- just to run this by you, because obviously I want to get something you'll like for upstream -- assuming you agree, is there a reason we don't coalesce the phys_avail array as well? I'm thinking that would be the simplest solution as it would restore the assumption in the segment walk loop.
The other obvious possibility is to try to adjust the end of the avail range being considered if the next avail range is adjacent. I've prototyped that, and got something that mostly works (an earlier small segment in Domain 0 still wasn't added right, and I'd need to chase that down), but figured it makes much more sense to ping you at this point and see what you'd be thinking.
Thanks again.

Dec 14 2020, 2:16 PM
dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

Mark -- first, apologies for going off on the wrong track on this originally. I got the idea of the order being wrong from other triage comments and ran with that. Mea culpa.

Dec 14 2020, 2:11 PM

Dec 10 2020

dgmorris_earthlink.net added a comment to D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..

which it isn't as that is at (SI_SUB_VM - 1) which is after CPU

I don't quite follow. SI_SUB_VM is before SI_SUB_CPU.

Dec 10 2020, 5:57 PM
dgmorris_earthlink.net requested review of D27546: Avoid vm_free_count() for physmem available before VM ACPI walk..
Dec 10 2020, 5:20 PM

Jul 16 2020

dgmorris_earthlink.net accepted D25689: Revert r240317 to prevent leaking pmap entries.

LGTM.

Jul 16 2020, 8:33 PM

Jul 3 2019

D20842: Bug 238960: panic in vm_pageout_collect_batch() with QUEUE_MACRO_DEBUG_TRASH enabled now requires review to proceed.

Moved new local variable n to appropriate spot in multiple declarations.

Jul 3 2019, 4:23 PM
dgmorris_earthlink.net created D20842: Bug 238960: panic in vm_pageout_collect_batch() with QUEUE_MACRO_DEBUG_TRASH enabled.
Jul 3 2019, 4:08 PM

May 9 2019

dgmorris_earthlink.net accepted D20219: ddb(4): Add 'sysctl' command.

Looks like a straightforward copy of what we reviewed, yeah.

May 9 2019, 11:51 PM

Aug 24 2018

dgmorris_earthlink.net updated the diff for D15119: Handle overflow of uid or gid in arguments for chown.

Updated per review comments -- did I miss anything?

Aug 24 2018, 2:50 PM · Contributor Reviews (base)

Apr 17 2018

dgmorris_earthlink.net updated the diff for D15119: Handle overflow of uid or gid in arguments for chown.

Uploaded file for diff instead of cut/paste that was introducing a whitespace issue.

Apr 17 2018, 8:34 PM · Contributor Reviews (base)
dgmorris_earthlink.net created D15119: Handle overflow of uid or gid in arguments for chown.
Apr 17 2018, 6:37 PM · Contributor Reviews (base)