Page MenuHomeFreeBSD
Feed Advanced Search

Fri, Oct 16

rlibby accepted D26772: uma: Respect uk_reserve in keg_drain().
Fri, Oct 16, 8:47 PM
rlibby accepted D26772: uma: Respect uk_reserve in keg_drain().

Looks good!

Fri, Oct 16, 4:48 PM

Thu, Oct 15

rlibby added inline comments to D26772: uma: Respect uk_reserve in keg_drain().
Thu, Oct 15, 5:37 PM
rlibby added inline comments to D26772: uma: Respect uk_reserve in keg_drain().
Thu, Oct 15, 8:04 AM

Wed, Oct 14

rlibby accepted D26771: uma: Avoid depleting keg reserves when filling a bucket.
Wed, Oct 14, 5:53 PM
rlibby accepted D26770: vmem: Allocate btags before looping in vmem_xalloc().

Change looks good.

Wed, Oct 14, 5:51 PM
rlibby accepted D26769: vmem: Simplify bt_fill() callers a bit.
Wed, Oct 14, 5:13 PM
rlibby added inline comments to D26771: uma: Avoid depleting keg reserves when filling a bucket.
Wed, Oct 14, 4:11 AM
rlibby accepted D26769: vmem: Simplify bt_fill() callers a bit.
Wed, Oct 14, 3:26 AM
rlibby added a comment to D26770: vmem: Allocate btags before looping in vmem_xalloc().

So, is it the case that if we do a successful vmem_import under vmem_try_fetch that that new segment must be able to satisfy the request (and we won't do a second etc import)? I notice that there are several arguments to vmem_fit which are not arguments to vmem_import.

Wed, Oct 14, 1:19 AM

Tue, Oct 13

rlibby accepted D26769: vmem: Simplify bt_fill() callers a bit.

Why not just put the logic in bt_fill(), is there a future patch that plans to use _bt_fill()?

Tue, Oct 13, 11:59 PM

Wed, Sep 30

rlibby accepted D26426: UMA: Use LIFO for non-SMR bucket caches..
Wed, Sep 30, 4:24 PM
rlibby added inline comments to D26426: UMA: Use LIFO for non-SMR bucket caches..
Wed, Sep 30, 4:40 AM
rlibby accepted D26599: gdb(4): Don't escape GDB special characters at application layer.

Seems fine, especially as we already squash other characters instead of
escaping them.

Wed, Sep 30, 4:21 AM

Sep 3 2020

rlibby abandoned D23937: amd64 atomics: guard flag output operand use with feature test.
Sep 3 2020, 6:16 PM

Sep 2 2020

rlibby added a comment to D26150: make m_getm2() resilient to zone_jumbop exhaustion.

Is there a possible side effect where if we now never do an M_WAITOK
alloc from zone_jumbop then we may not increase the zone size up to the
limit? As in, should there be a condition in here where we only convert
M_WAITOK to M_NOWAIT if we are actually near the limit? For example
should mb_reclaim arm this?

Sep 2 2020, 1:27 AM

Aug 18 2020

rlibby accepted D26115: gdb(4): Support empty qSupported queries.

LGTM.

Aug 18 2020, 8:02 PM

May 20 2020

rlibby added a comment to D24755: Permit deferred creation of SMR structures for the VM radix zone..

This makes me a little uneasy. Allocation before attach is okay, but free is not. Do we have some kind of guarantee that no free will occur?

May 20 2020, 9:02 PM

Mar 7 2020

rlibby updated subscribers of D23997: malloc debugging: detect mismatched mtp in alloc/free with redzone.
Mar 7 2020, 10:28 PM
rlibby updated the summary of D23997: malloc debugging: detect mismatched mtp in alloc/free with redzone.
Mar 7 2020, 10:12 PM
rlibby created D23997: malloc debugging: detect mismatched mtp in alloc/free with redzone.
Mar 7 2020, 10:10 PM
rlibby accepted D23980: Make uma_int.h friendlier to userspace..

Fixups LGTM.

Mar 7 2020, 12:52 AM

Mar 6 2020

rlibby accepted D23980: Make uma_int.h friendlier to userspace..

Looks good except SLAB_BITSETS and Jeff's comments.

Mar 6 2020, 10:20 PM
rlibby accepted D23988: Move SMR pointer type definition and access macros to smr_types.h..
Mar 6 2020, 8:02 PM

Mar 5 2020

rlibby added a comment to D23967: emulators/virtualbox-ose: use contemporary GCC instead of old llvm.

USE_GCC= any is not sufficient, as GCC8 doesn't support the feature used by atomic.h.

Mar 5 2020, 5:10 AM

Mar 3 2020

rlibby added a comment to D23937: amd64 atomics: guard flag output operand use with feature test.

Okay, I am happiest to go with option #2 also. I'll leave this up for a little in case any one else has a comment.

Mar 3 2020, 5:00 PM
rlibby created D23937: amd64 atomics: guard flag output operand use with feature test.
Mar 3 2020, 6:58 AM

Mar 1 2020

rlibby accepted D23902: vm: add debug to uma_zone_set_smr.

Looks good.

Mar 1 2020, 3:10 AM

Feb 29 2020

rlibby accepted D23895: Ensure that arm64 thread structures are allocated from the direct map..

Too bad. Are you planning to pursue this further or to move on?

Feb 29 2020, 6:07 PM
rlibby added a comment to D23661: Implement a translation from FreeBSD atomic API to gcc intrinsics.
In D23661#521387, @mjg wrote:

I think it's a fair worry that mixing hand-rolled atomics with compiler intrinsics may lead to trouble, but it's not hard to remedy.

Note the 2 mentioned missing testandclear and readandclear can be implemented without significant difficulty (the latter is a oneliner, the former worst case can be a cmpset loop. I grepped only 2 users.)

Feb 29 2020, 5:39 PM

Feb 28 2020

rlibby closed D23870: amd64 pmap.c: minor codegen optimization in flag access.
Feb 28 2020, 6:32 PM
rlibby committed rS358440: amd64 pmap.c: minor codegen optimization in flag access.
amd64 pmap.c: minor codegen optimization in flag access
Feb 28 2020, 6:32 PM
rlibby committed rS358439: amd64 atomic.h: minor codegen optimization in flag access.
amd64 atomic.h: minor codegen optimization in flag access
Feb 28 2020, 6:32 PM
rlibby closed D23869: amd64 atomic.h: minor codegen optimization in flag access.
Feb 28 2020, 6:32 PM
rlibby updated subscribers of D23870: amd64 pmap.c: minor codegen optimization in flag access.
Feb 28 2020, 4:57 PM
rlibby added a comment to D23869: amd64 atomic.h: minor codegen optimization in flag access.
In D23869#524905, @mjg wrote:

I only have a minor comment that the feature is somewhat new in clang, I don't know if we care [that is, if we should version check]. I also happened to have a complete rewrite here D23661, but I think this should go in regardless of what happens with that review.

Feb 28 2020, 4:49 PM
rlibby created D23870: amd64 pmap.c: minor codegen optimization in flag access.
Feb 28 2020, 5:55 AM
rlibby created D23869: amd64 atomic.h: minor codegen optimization in flag access.
Feb 28 2020, 5:50 AM

Feb 27 2020

rlibby accepted D23825: Simplify lazy advance and decouple it somewhat from ticks so that it canbe forwarded after other broadcast interrupts..

Looks good.

Feb 27 2020, 10:37 AM

Feb 25 2020

rlibby accepted D23824: A pair of UMA optimizations to improve cache behavior..

Looks good after fixing Mark's comments.

Feb 25 2020, 7:19 AM

Feb 24 2020

rlibby added a comment to D23802: vn_fullpath: const'ify retbuf.

I agree the interface is strange. It would seem simpler to me if it were just

int vn_fullpath(struct thread *td, struct vnode *vp, char **retbuf);

Feb 24 2020, 6:31 PM
rlibby added a comment to D23802: vn_fullpath: const'ify retbuf.

I think in terms of const semantics, this is okay, although it does seem accidental and undesirable to prevent callers from modifying the buf on success as a side-effect.

Feb 24 2020, 5:42 PM

Feb 23 2020

rlibby closed D23798: sys/kern: quiet -Wwrite-strings.
Feb 23 2020, 3:32 AM
rlibby committed rS358258: sys/kern: quiet -Wwrite-strings.
sys/kern: quiet -Wwrite-strings
Feb 23 2020, 3:32 AM
rlibby committed rS358257: vfs: quiet -Wwrite-strings.
vfs: quiet -Wwrite-strings
Feb 23 2020, 3:32 AM
rlibby closed D23796: sys/vm: quiet -Wwrite-strings.
Feb 23 2020, 3:32 AM
rlibby committed rS358256: sys/vm: quiet -Wwrite-strings.
sys/vm: quiet -Wwrite-strings
Feb 23 2020, 3:32 AM
rlibby closed D23797: vfs: quiet -Wwrite-strings.
Feb 23 2020, 3:32 AM

Feb 22 2020

rlibby created D23798: sys/kern: quiet -Wwrite-strings.
Feb 22 2020, 9:49 PM
rlibby created D23797: vfs: quiet -Wwrite-strings.
Feb 22 2020, 9:40 PM
rlibby created D23796: sys/vm: quiet -Wwrite-strings.
Feb 22 2020, 9:39 PM
rlibby added inline comments to D23533: Plumb a domain parameter through uma_reclaim() and callees..
Feb 22 2020, 4:57 PM
rlibby added inline comments to D23738: Experimental ticks based SMR..
Feb 22 2020, 6:46 AM

Feb 20 2020

rlibby added inline comments to D23738: Experimental ticks based SMR..
Feb 20 2020, 6:29 AM

Feb 19 2020

rlibby closed D23729: powerpc: unconditionally mark SLB zones UMA_ZONE_CONTIG.
Feb 19 2020, 4:47 AM
rlibby committed rS358094: powerpc: unconditionally mark SLB zones UMA_ZONE_CONTIG.
powerpc: unconditionally mark SLB zones UMA_ZONE_CONTIG
Feb 19 2020, 4:47 AM

Feb 17 2020

rlibby added a comment to D23729: powerpc: unconditionally mark SLB zones UMA_ZONE_CONTIG.

Can you explain why this makes a difference, at least for slbt_zone? I think I am missing something.

I can report observed behavior both ways:

Without removing the conditional structure the old PowerMac G5 "Quads"
do not boot. With the removal they do. The booting has been broken since
head -r357549 when the conditional allocf_flags code structure was put in.

See https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=244118 .

I'm now running with the patch from that bugzilla and have progressed to
-r357979 with it.

Feb 17 2020, 8:07 PM
rlibby added a comment to D23729: powerpc: unconditionally mark SLB zones UMA_ZONE_CONTIG.

Can you explain why this makes a difference, at least for slbt_zone? I think I am missing something.

Feb 17 2020, 7:53 PM
rlibby created D23729: powerpc: unconditionally mark SLB zones UMA_ZONE_CONTIG.
Feb 17 2020, 6:28 PM

Feb 7 2020

rlibby committed rS357644: smr.h: fix build after r357641.
smr.h: fix build after r357641
Feb 7 2020, 12:48 AM

Feb 6 2020

rlibby added a comment to D23464: Fix a race in smr_advance() that could result in unnecessary poll calls..
In D23464#516865, @jkim wrote:

It seems you forgot to define SMR_SEQ_DELTA() macro.

Feb 6 2020, 9:46 PM
rlibby accepted D23532: Optimize keg_drain()..

Fixups look good.

Feb 6 2020, 3:48 PM
rlibby committed rS357611: auditd_stop: wait_for_pids instead of sleeping.
auditd_stop: wait_for_pids instead of sleeping
Feb 6 2020, 8:32 AM
rlibby closed D23223: auditd_stop: wait_for_pids instead of sleeping.
Feb 6 2020, 8:32 AM
rlibby committed rS357610: uma: remove UMA_ZFLAG_CACHEONLY flag.
uma: remove UMA_ZFLAG_CACHEONLY flag
Feb 6 2020, 8:32 AM
rlibby closed D23516: uma: remove UMA_ZFLAG_CACHEONLY flag.
Feb 6 2020, 8:32 AM
rlibby accepted D23532: Optimize keg_drain()..
Feb 6 2020, 6:48 AM

Feb 5 2020

rlibby added inline comments to D22963: Expand generic subword atomic primitives.
Feb 5 2020, 7:24 PM
rlibby added inline comments to D22963: Expand generic subword atomic primitives.
Feb 5 2020, 5:32 PM
rlibby created D23516: uma: remove UMA_ZFLAG_CACHEONLY flag.
Feb 5 2020, 9:08 AM

Feb 4 2020

rlibby closed D23487: uma: multipage chicken switch.
Feb 4 2020, 10:40 PM
rlibby committed rS357550: uma: multipage chicken switch.
uma: multipage chicken switch
Feb 4 2020, 10:40 PM
rlibby committed rS357549: uma: grow slabs to enforce minimum memory efficiency.
uma: grow slabs to enforce minimum memory efficiency
Feb 4 2020, 10:40 PM
rlibby committed rS357548: uma: convert mbuf_jumbo_alloc to UMA_ZONE_CONTIG & tag others.
uma: convert mbuf_jumbo_alloc to UMA_ZONE_CONTIG & tag others
Feb 4 2020, 10:40 PM
rlibby closed D23239: [uma-multipage 3/3] uma: grow slabs to enforce minimum memory efficiency.
Feb 4 2020, 10:40 PM
rlibby closed D23238: [uma-multipage 2/3] uma: convert mbuf_jumbo_alloc & pmap_pdpt_alloc to UMA_ZONE_CONTIG.
Feb 4 2020, 10:40 PM
rlibby closed D23237: [uma-multipage 1/3] uma: add UMA_ZONE_CONTIG, and a default contig_alloc.
Feb 4 2020, 10:40 PM
rlibby committed rS357547: uma: add UMA_ZONE_CONTIG, and a default contig_alloc.
uma: add UMA_ZONE_CONTIG, and a default contig_alloc
Feb 4 2020, 10:40 PM
rlibby closed D23496: uma: pcpu_page_free needs to startup_free pages from startup_alloc.
Feb 4 2020, 10:40 PM
rlibby committed rS357546: uma: pcpu_page_free needs to startup_free pages from startup_alloc.
uma: pcpu_page_free needs to startup_free pages from startup_alloc
Feb 4 2020, 10:40 PM
rlibby added a comment to D23223: auditd_stop: wait_for_pids instead of sleeping.
In D23223#515942, @cem wrote:

I don't see any reason auditd cannot just use the default stop action, which sends SIGTERM. auditd handles sigterm and performs more or less the same shutdown actions as those triggered by audit -t (AUDIT_TRIGGER_CLOSE_AND_DIE). I would suggest just removing the auditd_stop() function entirely.

Feb 4 2020, 10:29 PM
rlibby added a comment to D23223: auditd_stop: wait_for_pids instead of sleeping.
In D23223#515931, @cem wrote:

Seems like there is a possible race condition here where $rc_pid exits and is recycled before we check for exit.

Feb 4 2020, 10:12 PM
rlibby updated the diff for D23487: uma: multipage chicken switch.

Rename from vm.multipage_slabs to vm.debug.uma_multipage_slabs

Feb 4 2020, 7:55 PM
rlibby updated the diff for D23239: [uma-multipage 3/3] uma: grow slabs to enforce minimum memory efficiency.

Fix CTR/KASSERT after dropping some local variables

Feb 4 2020, 7:52 PM
rlibby added a comment to D23238: [uma-multipage 2/3] uma: convert mbuf_jumbo_alloc & pmap_pdpt_alloc to UMA_ZONE_CONTIG.

Fix CTR/KASSERT after dropping some local variables

Feb 4 2020, 7:50 PM
rlibby updated the diff for D23238: [uma-multipage 2/3] uma: convert mbuf_jumbo_alloc & pmap_pdpt_alloc to UMA_ZONE_CONTIG.

Fix CTR/KASSERT after dropping some local variables

Feb 4 2020, 7:49 PM
rlibby updated the diff for D23237: [uma-multipage 1/3] uma: add UMA_ZONE_CONTIG, and a default contig_alloc.

Drop early-boot PCPU assert. It works now.

Feb 4 2020, 7:47 PM
rlibby added inline comments to D23487: uma: multipage chicken switch.
Feb 4 2020, 8:26 AM
rlibby updated subscribers of D23487: uma: multipage chicken switch.
Feb 4 2020, 8:25 AM
rlibby created D23496: uma: pcpu_page_free needs to startup_free pages from startup_alloc.
Feb 4 2020, 12:58 AM

Feb 3 2020

rlibby added inline comments to D23487: uma: multipage chicken switch.
Feb 3 2020, 10:32 PM
rlibby added inline comments to D23487: uma: multipage chicken switch.
Feb 3 2020, 6:37 PM
rlibby added inline comments to D23487: uma: multipage chicken switch.
Feb 3 2020, 6:22 PM
rlibby created D23487: uma: multipage chicken switch.
Feb 3 2020, 5:59 PM
rlibby updated the diff for D23237: [uma-multipage 1/3] uma: add UMA_ZONE_CONTIG, and a default contig_alloc.

Rebase over r357392.

Feb 3 2020, 5:49 PM
rlibby added inline comments to D23462: Deferred write seq advancement.
Feb 3 2020, 2:03 AM

Feb 2 2020

rlibby accepted D23464: Fix a race in smr_advance() that could result in unnecessary poll calls..

Looks good.

Feb 2 2020, 8:37 PM
rlibby accepted D23463: Add counters to smr..

Logic looks fine. I trust you've checked any performance effects.

Feb 2 2020, 8:23 PM

Jan 30 2020

rlibby added a comment to D22586: Implement safe memory reclamation in UMA..

Just dropping some nitpick comments before heading to bed. I looked through the interdiff for the latest update, but I was still intending to take another pass through the rest of the uma_core.c changes.

Jan 30 2020, 10:05 AM

Jan 27 2020

rlibby added inline comments to D22586: Implement safe memory reclamation in UMA..
Jan 27 2020, 6:08 PM