cse_cem_gmail_com (Conrad Meyer)
User

Projects

User does not belong to any projects.

User Details

User Since
Nov 10 2014, 9:28 PM (171 w, 2 d)

Recent Activity

Jul 17 2015

cse_cem_gmail_com updated subscribers of D1672: per-cpu page cache.
Jul 17 2015, 4:13 PM
cse_cem_gmail_com added a reviewer for D1672: per-cpu page cache: adrian.

What can I do to help this along?

Jul 17 2015, 4:12 PM

Jun 27 2015

cse_cem_gmail_com abandoned D2900: ffs_backgroundwritedone: Set INVAL on errored IOs to avoid bdirty.

Landed, with Konstantin's additional work (and Kirk's blessing) to avoid *CORRUPTION*, in r284887.

Jun 27 2015, 3:38 PM

Jun 26 2015

cse_cem_gmail_com added a comment to D2882: Reduce bqlock contention in binsfree.

Committed in r284719.

Jun 26 2015, 1:52 AM

Jun 25 2015

cse_cem_gmail_com updated the diff for D2900: ffs_backgroundwritedone: Set INVAL on errored IOs to avoid bdirty.

Address kib's feedback; add B_INVAL comment, add failpoint.

Jun 25 2015, 12:28 AM

Jun 24 2015

cse_cem_gmail_com added a comment to D2900: ffs_backgroundwritedone: Set INVAL on errored IOs to avoid bdirty.
In D2900#56463, @kib wrote:

I would add a comment explaining the reason for B_INVAL there.

Jun 24 2015, 7:05 PM
cse_cem_gmail_com retitled D2900: ffs_backgroundwritedone: Set INVAL on errored IOs to avoid bdirty from to ffs_backgroundwritedone: Set INVAL on errored IOs to avoid bdirty.
Jun 24 2015, 5:04 PM
cse_cem_gmail_com abandoned D2883: bxe_free_fp_buffers: Take TX lock to access tx_br..

Fixed in r284739, discarding this review.

Jun 24 2015, 3:03 PM

Jun 22 2015

cse_cem_gmail_com updated the diff for D2882: Reduce bqlock contention in binsfree.

Move olock calculation to only where it is needed.

Jun 22 2015, 6:24 PM
cse_cem_gmail_com planned changes to D2882: Reduce bqlock contention in binsfree.

Will fix.

Jun 22 2015, 6:21 PM
cse_cem_gmail_com added a comment to D2883: bxe_free_fp_buffers: Take TX lock to access tx_br..
In D2883#55943, @benno wrote:

This should be looked at by davidcs@. I'm going to email him to see if I can get him signed up for Phabricator.

Jun 22 2015, 5:26 PM
cse_cem_gmail_com added inline comments to D2883: bxe_free_fp_buffers: Take TX lock to access tx_br..
Jun 22 2015, 5:19 PM
cse_cem_gmail_com retitled D2883: bxe_free_fp_buffers: Take TX lock to access tx_br. from to bxe_free_fp_buffers: Take TX lock to access tx_br..
Jun 22 2015, 5:05 PM
cse_cem_gmail_com retitled D2882: Reduce bqlock contention in binsfree from to Reduce bqlock contention in binsfree.
Jun 22 2015, 4:42 PM

Jun 20 2015

cse_cem_gmail_com added a comment to D2852: Fix bug 200942 - Don't squash fts_children errors.

Thanks. Can you commit this for me? I don't have a src bit yet.

Jun 20 2015, 3:50 PM

Jun 19 2015

cse_cem_gmail_com updated the diff for D2852: Fix bug 200942 - Don't squash fts_children errors.

Address the concern I brought up re: using stale errno value after fchdir(2).

Jun 19 2015, 10:41 PM
cse_cem_gmail_com added inline comments to D2852: Fix bug 200942 - Don't squash fts_children errors.
Jun 19 2015, 10:27 PM
cse_cem_gmail_com added a comment to D2852: Fix bug 200942 - Don't squash fts_children errors.

Thanks!

Jun 19 2015, 10:22 PM
cse_cem_gmail_com updated the diff for D2852: Fix bug 200942 - Don't squash fts_children errors.

Per Jilles' feedback, only set serrno value if fts_build actually fails.

Jun 19 2015, 10:22 PM
cse_cem_gmail_com updated the test plan for D2852: Fix bug 200942 - Don't squash fts_children errors.
Jun 19 2015, 12:28 AM

Jun 17 2015

cse_cem_gmail_com updated the diff for D2852: Fix bug 200942 - Don't squash fts_children errors.

Don't preserve errno from close(2)

Jun 17 2015, 11:39 PM
cse_cem_gmail_com retitled D2852: Fix bug 200942 - Don't squash fts_children errors from to Fix bug 200942 - Don't squash fts_children errors.
Jun 17 2015, 10:58 PM

Jun 11 2015

cse_cem_gmail_com retitled D2181: vt: Draw logos per CPU core from vt: Draw beasties per CPU core to vt: Draw logos per CPU core.
Jun 11 2015, 1:33 PM

Jun 10 2015

cse_cem_gmail_com updated the diff for D2181: vt: Draw logos per CPU core.

Shuffle most new code to a separate file under dev/vt/; mostly use the word 'logo' in place of the word 'beastie' when referring to one of these things.

Jun 10 2015, 8:08 PM

May 23 2015

cse_cem_gmail_com closed D2612: Fix thread suspend deadlock by minding TDF_SBDRY.

Committed as r283320. Thanks Konstantin, Peter.

May 23 2015, 8:16 PM

May 22 2015

cse_cem_gmail_com updated subscribers of D2612: Fix thread suspend deadlock by minding TDF_SBDRY.
May 22 2015, 11:05 AM
cse_cem_gmail_com updated D2612: Fix thread suspend deadlock by minding TDF_SBDRY.
May 22 2015, 12:05 AM
cse_cem_gmail_com updated D2612: Fix thread suspend deadlock by minding TDF_SBDRY.
May 22 2015, 12:05 AM
cse_cem_gmail_com updated D2612: Fix thread suspend deadlock by minding TDF_SBDRY.
May 22 2015, 12:03 AM
cse_cem_gmail_com retitled D2612: Fix thread suspend deadlock by minding TDF_SBDRY from to Fix thread suspend deadlock by minding TDF_SBDRY.
May 22 2015, 12:01 AM

May 20 2015

cse_cem_gmail_com abandoned D2602: imgact_elf: Make note_procstat_files best-effort to avoid deadlock.

My mistake. Sleeping with locks held in a way that thread_single() can interrupt is invalid. Thanks, kib@. Sorry for the spam, everyone.

May 20 2015, 8:54 PM
cse_cem_gmail_com retitled D2602: imgact_elf: Make note_procstat_files best-effort to avoid deadlock from to imgact_elf: Make note_procstat_files best-effort to avoid deadlock.
May 20 2015, 8:41 PM

May 6 2015

cse_cem_gmail_com added a comment to D2460: Add SLIT enumeration and memory locality table to VM/ACPI..

Looks ok to me. There are some uncodified assumptions in the ACPI to vm_locality mapping that I'd prefer to see in some sort of ASSERT form. The printfs in the parsing probably go away or are cleaned up too.

May 6 2015, 1:43 AM

Apr 29 2015

cse_cem_gmail_com added a comment to D2181: vt: Draw logos per CPU core.

This feature is really cool!

Apr 29 2015, 12:50 PM

Apr 28 2015

cse_cem_gmail_com closed D2381: vm_fault: Fix deadlock between vm_map_wire and vm_fault_hold.

Committed by @kostikbel in r282128.

Apr 28 2015, 11:58 PM
cse_cem_gmail_com abandoned D2250: Reduce CPU burn on secondary CPUs during panic.

This breaks reboot-from-panic when the panic CPU is not the BSP CPU. Discarding.

Apr 28 2015, 9:21 PM

Apr 27 2015

cse_cem_gmail_com retitled D2381: vm_fault: Fix deadlock between vm_map_wire and vm_fault_hold from to vm_fault: Fix deadlock between vm_map_wire and vm_fault_hold.
Apr 27 2015, 10:05 PM

Apr 16 2015

cse_cem_gmail_com added a comment to D2181: vt: Draw logos per CPU core.
In D2181#12, @adrian wrote:

This is a really slow way of drawing things.

Apr 16 2015, 6:32 PM
cse_cem_gmail_com added a comment to D2181: vt: Draw logos per CPU core.
In D2181#10, @ed wrote:

I think one of the important things to do is to keep vt(4) simple. syscons(4) became so complex that nobody really understood how it worked anymore. I am afraid that patches that add gimmicks like these negatively affect the overall readability of vt(4) for little functional gain.

Apr 16 2015, 5:56 PM

Apr 9 2015

cse_cem_gmail_com updated subscribers of D2271: defer inactive processing (re: D1677).
Apr 9 2015, 11:02 PM
cse_cem_gmail_com added a comment to D2181: vt: Draw logos per CPU core.

Ping. What can I do to help move this forward?

Apr 9 2015, 1:57 PM
cse_cem_gmail_com updated the diff for D2250: Reduce CPU burn on secondary CPUs during panic.

Drop disable_intr(); we're in NMI context and interrupts are disabled.

Apr 9 2015, 1:50 PM

Apr 8 2015

cse_cem_gmail_com added a comment to D2250: Reduce CPU burn on secondary CPUs during panic.
In D2250#5, @kostikbel wrote:

This means that CPU cannot be directed to make any further work.

Apr 8 2015, 2:00 AM

Apr 7 2015

cse_cem_gmail_com added inline comments to D2250: Reduce CPU burn on secondary CPUs during panic.
Apr 7 2015, 6:57 PM
cse_cem_gmail_com retitled D2250: Reduce CPU burn on secondary CPUs during panic from to Reduce CPU burn on secondary CPUs during panic.
Apr 7 2015, 6:24 PM
cse_cem_gmail_com added a reviewer for D1895: buf_flags: Add missing B_KVAALLOC, B_UNMAPPED: markj.
Apr 7 2015, 5:53 PM

Apr 6 2015

cse_cem_gmail_com closed D2246: lockf.1: Mention lockf(3).

r281175

Apr 6 2015, 11:03 PM
cse_cem_gmail_com accepted D2246: lockf.1: Mention lockf(3).

r281175

Apr 6 2015, 11:03 PM
cse_cem_gmail_com retitled D2246: lockf.1: Mention lockf(3) from to lockf.1: Mention lockf(3).
Apr 6 2015, 9:47 PM

Apr 5 2015

cse_cem_gmail_com updated subscribers of D323: Allow stack(9) to capture stack of running process via stack_save_thread(). Have kern.proc.kstack use this..
In D323#5, @jhb wrote:

The in_panic change looks a bit interesting, but I'm curious if that is actually used. A userland process can't be scheduled to run to even invoke this sysctl if the system is panic'ing. Does Isilon use stack_save_thread() in other places?

Apr 5 2015, 3:05 AM

Apr 2 2015

cse_cem_gmail_com closed D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).
Apr 2 2015, 8:29 PM
cse_cem_gmail_com accepted D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Committed by kib@ in r281003. Thanks!

Apr 2 2015, 8:29 PM
cse_cem_gmail_com updated the diff for D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Move cleanup to the end of relocation; preloaded and runtime-loaded modules both go through this path.

Apr 2 2015, 12:07 PM
cse_cem_gmail_com added a comment to D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Damnit, wrong patch. Just a sec.

Apr 2 2015, 12:05 PM
cse_cem_gmail_com updated the diff for D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Move cleanup to the end of relocation; preloaded and runtime-loaded modules both go through this path.

Apr 2 2015, 12:05 PM

Apr 1 2015

cse_cem_gmail_com updated the diff for D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Updated to revert the SHN_FBSD_CACHED symtable changes after loading has completed, per (2) above.

Apr 1 2015, 4:31 PM

Mar 31 2015

cse_cem_gmail_com added a comment to D2181: vt: Draw logos per CPU core.
In D2181#3, @emaste wrote:

How do you handle ncpu greater than the number of Beasties that fit on a row?

Mar 31 2015, 2:11 PM
cse_cem_gmail_com added a comment to D2181: vt: Draw logos per CPU core.
In D2181#3, @emaste wrote:

How do you handle ncpu greater than the number of Beasties that fit on a row?

Mar 31 2015, 2:03 PM
cse_cem_gmail_com retitled D2181: vt: Draw logos per CPU core from to vt: Draw beasties per CPU core.
Mar 31 2015, 1:51 PM

Mar 21 2015

cse_cem_gmail_com added a comment to D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).
In D1718#13, @kostikbel wrote:

Ok. There are two possibilities:

  1. Introduce symbol cache used during the module load, and discarded afterward. You can use hash for the symbol names there, which would not help for the first lookup, but does provide more speedup for the duplicate lookups (more than your current patch).
Mar 21 2015, 3:26 PM
cse_cem_gmail_com added a comment to D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).
In D1718#11, @kostikbel wrote:

What bug does the patch fix ?

Mar 21 2015, 2:49 PM
cse_cem_gmail_com added a comment to D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).
In D1718#9, @kostikbel wrote:

This is hack.

Mar 21 2015, 1:23 AM

Feb 27 2015

cse_cem_gmail_com closed D1980: Relax vflush FORCECLOSE special vnode to only protect devfs vnodes.
Feb 27 2015, 4:45 PM
cse_cem_gmail_com accepted D1980: Relax vflush FORCECLOSE special vnode to only protect devfs vnodes.

Committed in r279362.

Feb 27 2015, 4:45 PM
cse_cem_gmail_com updated the diff for D1980: Relax vflush FORCECLOSE special vnode to only protect devfs vnodes.

Per discussion on IRC. Move extern declaration into sys/vnode.h so we don't need to import devfs.h in vfs_subr.c.

Feb 27 2015, 4:26 PM
cse_cem_gmail_com retitled D1980: Relax vflush FORCECLOSE special vnode to only protect devfs vnodes from to Relax vflush FORCECLOSE special vnode to only protect devfs vnodes.
Feb 27 2015, 4:18 PM

Feb 18 2015

cse_cem_gmail_com retitled D1895: buf_flags: Add missing B_KVAALLOC, B_UNMAPPED from to buf_flags: Add missing B_KVAALLOC, B_UNMAPPED.
Feb 18 2015, 4:37 PM
cse_cem_gmail_com added a comment to D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Ping! Anything I can do to help further this along?

Feb 18 2015, 4:36 PM
cse_cem_gmail_com closed D1724: rtld-elf should use powerof2() (PR 192238).
Feb 18 2015, 4:36 PM
cse_cem_gmail_com accepted D1724: rtld-elf should use powerof2() (PR 192238).
Feb 18 2015, 4:36 PM
cse_cem_gmail_com closed D1688: db_show_mount: Add mnt_lockref to ddb 'show mount'.

Closing as committed.

Feb 18 2015, 4:35 PM

Feb 17 2015

cse_cem_gmail_com added a comment to D1688: db_show_mount: Add mnt_lockref to ddb 'show mount'.

Thanks!

Feb 17 2015, 12:33 PM
cse_cem_gmail_com added a comment to D1688: db_show_mount: Add mnt_lockref to ddb 'show mount'.

Ping! Can someone take and commit this? I don't have a bit and phabricator thinks I should do it.

Feb 17 2015, 4:22 AM

Jan 30 2015

cse_cem_gmail_com retitled D1724: rtld-elf should use powerof2() (PR 192238) from rtld-elf should use powerof2(); nmaskwords must not be zero (PR 192238) to rtld-elf should use powerof2() (PR 192238).
Jan 30 2015, 2:56 PM
cse_cem_gmail_com updated the diff for D1724: rtld-elf should use powerof2() (PR 192238).

Drop nmaskwords > 0; remove superfluous nmw_power2 variable.

Jan 30 2015, 2:56 PM

Jan 29 2015

cse_cem_gmail_com updated the diff for D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Add 9999 context, drop fbt_powerpc.c modification (it has since been deleted or merged with fbt.c).

Jan 29 2015, 6:41 PM
cse_cem_gmail_com added a comment to D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).

Sorry, forgot on this one. I'll regenerate.

Jan 29 2015, 6:37 PM
cse_cem_gmail_com retitled D1724: rtld-elf should use powerof2() (PR 192238) from to rtld-elf should use powerof2(); nmaskwords must not be zero (PR 192238).
Jan 29 2015, 5:53 PM
cse_cem_gmail_com retitled D1718: Load reloc modules (amd64, mips) faster by caching global syms (PR 192249) from to Load reloc modules (amd64, mips) faster by caching global syms (PR 192249).
Jan 29 2015, 3:25 PM

Jan 27 2015

cse_cem_gmail_com retitled D1688: db_show_mount: Add mnt_lockref to ddb 'show mount' from to db_show_mount: Add mnt_lockref to ddb 'show mount'.
Jan 27 2015, 2:23 PM
cse_cem_gmail_com closed D1684: Fix vfs_busy leaks in ufs_quotactl.

Thanks!

Jan 27 2015, 1:07 PM
cse_cem_gmail_com added a reviewer for D1684: Fix vfs_busy leaks in ufs_quotactl: mckusick.
Jan 27 2015, 12:20 AM
cse_cem_gmail_com retitled D1684: Fix vfs_busy leaks in ufs_quotactl from to Fix vfs_busy leaks in ufs_quotactl.
Jan 27 2015, 12:17 AM

Jan 17 2015

cse_cem_gmail_com added a comment to D1548: Add support for parsing UCL config files to newsyslog.

LGTM.

Jan 17 2015, 4:41 PM
cse_cem_gmail_com added a comment to D1548: Add support for parsing UCL config files to newsyslog.

LGTM.

Jan 17 2015, 2:18 PM
cse_cem_gmail_com added a comment to D1548: Add support for parsing UCL config files to newsyslog.

Looks pretty good!

Jan 17 2015, 3:35 AM
cse_cem_gmail_com added a comment to D1548: Add support for parsing UCL config files to newsyslog.

Changes between 3218 and 3220 LGTM.

Jan 17 2015, 3:05 AM
cse_cem_gmail_com added inline comments to D1548: Add support for parsing UCL config files to newsyslog.
Jan 17 2015, 3:00 AM
cse_cem_gmail_com updated subscribers of D1548: Add support for parsing UCL config files to newsyslog.

Not a super thorough review, just passed over briefly. I'm not an expert on UCL nor newsyslog.

Jan 17 2015, 1:57 AM

Dec 13 2014

cse_cem_gmail_com added a comment to D904: Refactor dumpsys() to remove excessive code duplication.

Most of the changes are mechanical, so I can do them myself if you like.

Dec 13 2014, 12:42 AM

Dec 5 2014

cse_cem_gmail_com abandoned D1263: rtld-elf: Unmap unused object segments.
Dec 5 2014, 2:31 PM

Dec 3 2014

cse_cem_gmail_com added a comment to D1263: rtld-elf: Unmap unused object segments.
In D1263#8, @kostikbel wrote:

Can you enumerate the issues ?

Dec 3 2014, 8:53 PM
cse_cem_gmail_com added a comment to D1263: rtld-elf: Unmap unused object segments.

In short, mappable holes between segments tends to be reused, which causes weird issues.

Dec 3 2014, 8:26 PM
cse_cem_gmail_com edited reviewers for D1263: rtld-elf: Unmap unused object segments, added: kib; removed: kibab.
Dec 3 2014, 7:59 PM
cse_cem_gmail_com retitled D1263: rtld-elf: Unmap unused object segments from to rtld-elf: Unmap unused object segments.
Dec 3 2014, 7:55 PM

Dec 2 2014

cse_cem_gmail_com accepted D1152: (PR 194296) bfd: Improve common ELF linking error message.
Dec 2 2014, 12:59 AM

Dec 1 2014

cse_cem_gmail_com accepted D1152: (PR 194296) bfd: Improve common ELF linking error message.

Er, mark reviewed. Phabricator is hard.

Dec 1 2014, 11:03 PM
cse_cem_gmail_com added a comment to D1152: (PR 194296) bfd: Improve common ELF linking error message.

If you've tested it and it works, looks good to me. I just want this to be fixed, and every iteration seems like a basically adequate way of solving the problem to me.

Dec 1 2014, 11:02 PM

Nov 19 2014

cse_cem_gmail_com added inline comments to D904: Refactor dumpsys() to remove excessive code duplication.
Nov 19 2014, 4:50 AM

Nov 12 2014

cse_cem_gmail_com accepted D1152: (PR 194296) bfd: Improve common ELF linking error message.

Looks good to me.

Nov 12 2014, 8:28 PM