Page MenuHomeFreeBSD

mav (Alexander Motin)
User

Projects

User Details

User Since
Jun 3 2014, 6:27 PM (249 w, 6 d)

Recent Activity

Fri, Mar 15

mav committed rS345200: MFV r336930: 9284 arc_reclaim_thread has 2 jobs.
MFV r336930: 9284 arc_reclaim_thread has 2 jobs
Fri, Mar 15, 6:59 PM

Thu, Mar 14

mav committed rS345123: MFC r344903: Improve entropy for ZFS taskqueue selection..
MFC r344903: Improve entropy for ZFS taskqueue selection.
Thu, Mar 14, 12:59 AM
mav committed rS345122: MFC r344903: Improve entropy for ZFS taskqueue selection..
MFC r344903: Improve entropy for ZFS taskqueue selection.
Thu, Mar 14, 12:59 AM
mav committed rS345121: MFC r344866: Add respective tunables to few ZFS sysctls..
MFC r344866: Add respective tunables to few ZFS sysctls.
Thu, Mar 14, 12:58 AM
mav committed rS345120: MFC r344866: Add respective tunables to few ZFS sysctls..
MFC r344866: Add respective tunables to few ZFS sysctls.
Thu, Mar 14, 12:57 AM

Wed, Mar 13

mav committed rS345116: MFC r344844: Flush stdout after each iteration..
MFC r344844: Flush stdout after each iteration.
Wed, Mar 13, 8:29 PM
mav committed rS345115: MFC r344844: Flush stdout after each iteration..
MFC r344844: Flush stdout after each iteration.
Wed, Mar 13, 8:29 PM
mav committed rS345114: MFC r344636: Refactor command ordering/blocking mechanism in CTL..
MFC r344636: Refactor command ordering/blocking mechanism in CTL.
Wed, Mar 13, 8:28 PM
mav committed rS345113: MFC r344636: Refactor command ordering/blocking mechanism in CTL..
MFC r344636: Refactor command ordering/blocking mechanism in CTL.
Wed, Mar 13, 8:28 PM
mav committed rS345112: MFC r344586: Scrap some debug printf's, unused for years..
MFC r344586: Scrap some debug printf's, unused for years.
Wed, Mar 13, 8:26 PM
mav committed rS345111: MFC r344586: Scrap some debug printf's, unused for years..
MFC r344586: Scrap some debug printf's, unused for years.
Wed, Mar 13, 8:26 PM

Tue, Mar 12

mav committed rS345069: MFC r339299: Pull in a follow-on commit to resolve a deadlock in ZFS.
MFC r339299: Pull in a follow-on commit to resolve a deadlock in ZFS
Tue, Mar 12, 4:46 PM

Mon, Mar 11

mav committed rS345014: Revert minor part of r344934..
Revert minor part of r344934.
Mon, Mar 11, 5:39 PM
mav committed rS345007: MFC r344743: Reduce CTL threads priority to about PUSER..
MFC r344743: Reduce CTL threads priority to about PUSER.
Mon, Mar 11, 1:57 PM
mav committed rS345006: MFC r344743: Reduce CTL threads priority to about PUSER..
MFC r344743: Reduce CTL threads priority to about PUSER.
Mon, Mar 11, 1:56 PM
mav committed rS344993: MFC r344782: bridge: Fix spurious warnings about capabilities.
MFC r344782: bridge: Fix spurious warnings about capabilities
Mon, Mar 11, 1:45 AM
mav committed rS344992: MFC r344489: Free some space in struct ctl_io_hdr for better use..
MFC r344489: Free some space in struct ctl_io_hdr for better use.
Mon, Mar 11, 1:44 AM
mav committed rS344991: MFC r344489: Free some space in struct ctl_io_hdr for better use..
MFC r344489: Free some space in struct ctl_io_hdr for better use.
Mon, Mar 11, 1:44 AM

Fri, Mar 8

mav committed rS344936: MFV/ZoL: Disable LBA weighting on files and SSDs.
MFV/ZoL: Disable LBA weighting on files and SSDs
Fri, Mar 8, 9:13 PM
mav committed rS344934: Add separate aggregation limit for non-rotating media..
Add separate aggregation limit for non-rotating media.
Fri, Mar 8, 7:39 PM
mav committed rS344926: MFV/ZoL: Fix zfs_vdev_aggregation_limit bounds checking.
MFV/ZoL: Fix zfs_vdev_aggregation_limit bounds checking
Fri, Mar 8, 6:49 PM
mav committed rS344911: MFC r344661, r344669: Limit 24xx adapters to only MSI interrupts by default..
MFC r344661, r344669: Limit 24xx adapters to only MSI interrupts by default.
Fri, Mar 8, 12:56 AM
mav committed rS344910: MFC r344661, r344669: Limit 24xx adapters to only MSI interrupts by default..
MFC r344661, r344669: Limit 24xx adapters to only MSI interrupts by default.
Fri, Mar 8, 12:56 AM
mav committed rS344909: MFC r344660: Add to isp(4) tunables to limit MSI/MSI-X usage..
MFC r344660: Add to isp(4) tunables to limit MSI/MSI-X usage.
Fri, Mar 8, 12:55 AM
mav committed rS344908: MFC r344660: Add to isp(4) tunables to limit MSI/MSI-X usage..
MFC r344660: Add to isp(4) tunables to limit MSI/MSI-X usage.
Fri, Mar 8, 12:54 AM
mav committed rS344907: MFC r344702: There is no `device atacard` but there is `device atapccard`..
MFC r344702: There is no `device atacard` but there is `device atapccard`.
Fri, Mar 8, 12:53 AM
mav committed rS344906: MFC r344702: There is no `device atacard` but there is `device atapccard`..
MFC r344702: There is no `device atacard` but there is `device atapccard`.
Fri, Mar 8, 12:53 AM

Thu, Mar 7

mav committed rS344903: Improve entropy for ZFS taskqueue selection..
Improve entropy for ZFS taskqueue selection.
Thu, Mar 7, 10:56 PM
mav committed rS344866: Add respective tunables to few ZFS sysctls..
Add respective tunables to few ZFS sysctls.
Thu, Mar 7, 1:24 AM

Wed, Mar 6

mav committed rS344844: Flush stdout after each iteration..
Flush stdout after each iteration.
Wed, Mar 6, 3:19 PM

Mon, Mar 4

mav closed D18924: bridge: Fix spurious warnings about capabilities.
Mon, Mar 4, 10:01 PM
mav committed rS344782: bridge: Fix spurious warnings about capabilities.
bridge: Fix spurious warnings about capabilities
Mon, Mar 4, 10:01 PM
mav committed rS344743: Reduce CTL threads priority to about PUSER..
Reduce CTL threads priority to about PUSER.
Mon, Mar 4, 12:49 AM

Sun, Mar 3

mav added a comment to D19369: Make isp(4) suggest loading ispfw(4) when it fails to attach..

Regarding the message - why? I quite like the current one, because it tells you exactly what you need to do, instead of having to figure it out.

Sun, Mar 3, 6:53 PM

Sat, Mar 2

mav committed rS344717: MFC r344480: Remove disabled CTL_LEGACY_STATS support..
MFC r344480: Remove disabled CTL_LEGACY_STATS support.
Sat, Mar 2, 3:17 AM

Fri, Mar 1

mav committed rS344702: There is no `device atacard` but there is `device atapccard`..
There is no `device atacard` but there is `device atapccard`.
Fri, Mar 1, 3:00 PM
mav added a comment to D19408: camcontrol: possibly wrong sector count in ata_do_identify().

I agree, zero should be the right answer here. That is what kernel does also.

Fri, Mar 1, 2:38 PM

Thu, Feb 28

mav committed rS344669: Limit 24xx adapters to only MSI interrupts by default..
Limit 24xx adapters to only MSI interrupts by default.
Thu, Feb 28, 9:07 PM
mav committed rS344661: Limit 24xx adapters to only one MSI-X interrupt by default..
Limit 24xx adapters to only one MSI-X interrupt by default.
Thu, Feb 28, 3:36 PM
mav committed rS344660: Add to isp(4) tunables to limit MSI/MSI-X usage..
Add to isp(4) tunables to limit MSI/MSI-X usage.
Thu, Feb 28, 3:24 PM
mav accepted D19017: - Fix Headset sound for DELL laptops.

I have no objections, but I have no documentation to know what that coefficient being set means.

Thu, Feb 28, 1:34 PM
mav committed rS344652: MFC r344437: Allow I/OAT of present Xeon E5/E7 to work thorugh PLX NTB..
MFC r344437: Allow I/OAT of present Xeon E5/E7 to work thorugh PLX NTB.
Thu, Feb 28, 12:30 AM
mav committed rS344651: MFC r344437: Allow I/OAT of present Xeon E5/E7 to work thorugh PLX NTB..
MFC r344437: Allow I/OAT of present Xeon E5/E7 to work thorugh PLX NTB.
Thu, Feb 28, 12:30 AM
mav committed rS344650: MFC r344441: Fix few issues in ioat(4) driver..
MFC r344441: Fix few issues in ioat(4) driver.
Thu, Feb 28, 12:29 AM
mav committed rS344649: MFC r344441: Fix few issues in ioat(4) driver..
MFC r344441: Fix few issues in ioat(4) driver.
Thu, Feb 28, 12:28 AM

Wed, Feb 27

mav committed rS344636: Refactor command ordering/blocking mechanism in CTL..
Refactor command ordering/blocking mechanism in CTL.
Wed, Feb 27, 9:29 PM

Tue, Feb 26

mav added a comment to D19369: Make isp(4) suggest loading ispfw(4) when it fails to attach..

I think generally it may be not a bad idea, but I'd prefer less exact message, like "Consider loading ispfw kernel module for known good firmware".

Tue, Feb 26, 8:35 PM
mav committed rS344586: Scrap some debug printf's, unused for years..
Scrap some debug printf's, unused for years.
Tue, Feb 26, 4:05 PM
mav accepted D19184: Set process title during zfs send.

I have no more objections.

Tue, Feb 26, 4:23 AM

Mon, Feb 25

mav added a reviewer for D19184: Set process title during zfs send: ZFS.
Mon, Feb 25, 7:26 PM

Sat, Feb 23

mav committed rS344489: Free some space in struct ctl_io_hdr for better use..
Free some space in struct ctl_io_hdr for better use.
Sat, Feb 23, 11:35 PM
mav committed rS344480: Remove disabled CTL_LEGACY_STATS support..
Remove disabled CTL_LEGACY_STATS support.
Sat, Feb 23, 4:24 AM

Thu, Feb 21

mav committed rS344441: Fix few issues in ioat(4) driver..
Fix few issues in ioat(4) driver.
Thu, Feb 21, 4:47 PM
mav closed D19231: Fix few issues in ioat(4) driver..
Thu, Feb 21, 4:47 PM
mav committed rS344437: Allow I/OAT of present Xeon E5/E7 to work thorugh PLX NTB..
Allow I/OAT of present Xeon E5/E7 to work thorugh PLX NTB.
Thu, Feb 21, 2:10 PM
mav added a comment to D19231: Fix few issues in ioat(4) driver..

Any more comments?

Thu, Feb 21, 3:40 AM
mav committed rS344401: MFC r302669,302677-302686,303761,304602,304603,305027-305028,305259,.
MFC r302669,302677-302686,303761,304602,304603,305027-305028,305259,
Thu, Feb 21, 12:44 AM
mav committed rS344400: MFC r343125 (by cem): ioat(4): Set __result_use_check on ioat_acquire_reserve.
MFC r343125 (by cem): ioat(4): Set __result_use_check on ioat_acquire_reserve
Thu, Feb 21, 12:17 AM

Wed, Feb 20

mav updated the diff for D19231: Fix few issues in ioat(4) driver..

Fix few things noticed by cem@.

Wed, Feb 20, 2:48 AM
mav added inline comments to D19231: Fix few issues in ioat(4) driver..
Wed, Feb 20, 2:29 AM
mav added a comment to D19231: Fix few issues in ioat(4) driver..
In D19231#412056, @cem wrote:

How much testing have you been able to perform with this patch?

Wed, Feb 20, 2:28 AM

Mon, Feb 18

mav created D19231: Fix few issues in ioat(4) driver..
Mon, Feb 18, 1:25 AM

Feb 13 2019

mav added a comment to D19094: Restore ARC MFU/MRU pressure.

Don't sure about calling remove_reference() from arc_hdr_alloc_pabd() (or from parallel tasks), but see at ARC MFU/MRU size calculation in arc_change_state() called from arc_access() and !GHOST_STATE(state) case in arc_get_data_impl() called from arc_hdr_alloc_pabd().
I mean interchange this lines can cause problems for this accountings.

Feb 13 2019, 5:53 PM · ZFS
mav committed rS344076: MFC r343586: Remove BIO_ORDERED flag from BIO_FLUSH sent by ZFS..
MFC r343586: Remove BIO_ORDERED flag from BIO_FLUSH sent by ZFS.
Feb 13 2019, 12:39 AM
mav committed rS344075: MFC r343586: Remove BIO_ORDERED flag from BIO_FLUSH sent by ZFS..
MFC r343586: Remove BIO_ORDERED flag from BIO_FLUSH sent by ZFS.
Feb 13 2019, 12:39 AM
mav committed rS344074: MFC r343582,r343588:Relax BIO_FLUSH ordering in da(4), respecting BIO_ORDERED..
MFC r343582,r343588:Relax BIO_FLUSH ordering in da(4), respecting BIO_ORDERED.
Feb 13 2019, 12:38 AM
mav committed rS344073: MFC r343582,r343588:Relax BIO_FLUSH ordering in da(4), respecting BIO_ORDERED..
MFC r343582,r343588:Relax BIO_FLUSH ordering in da(4), respecting BIO_ORDERED.
Feb 13 2019, 12:37 AM
mav committed rS344072: MFC r343585: Only sort requests of types that have concept of offset..
MFC r343585: Only sort requests of types that have concept of offset.
Feb 13 2019, 12:35 AM
mav committed rS344071: MFC r343585: Only sort requests of types that have concept of offset..
MFC r343585: Only sort requests of types that have concept of offset.
Feb 13 2019, 12:33 AM

Feb 12 2019

mav added a comment to D19094: Restore ARC MFU/MRU pressure.
In D19094#410076, @lev wrote:

Huh? As far as I can see, problem not remove_reference(), but arc_adapt() called too late, with header in wrong state (promoted from something-ghost to live LRU), which brraks main idea of ARC adaptation.

Feb 12 2019, 11:01 PM · ZFS
mav added a comment to D19094: Restore ARC MFU/MRU pressure.

While I see the problem you are fixing, the fix looks ugly to me, that is why I would look for something nicer. I agree that according to logic of remove_reference() dropping last reference for header in ghost state is a failure, but how can remove_reference() be called before the arc_access() just on following line? I would guess from description telling about the case of prefetch read it should happen no sooner then we actually initiate the I/O, which is done much later then those two lines. So while I agree it is somewhat odd to have buffer for header in ghost state, is that a criminal.

Feb 12 2019, 9:54 PM · ZFS
mav committed rS344033: MFC r343562, r343563: Reimplement BIO_ORDERED handling in nvd(4)..
MFC r343562, r343563: Reimplement BIO_ORDERED handling in nvd(4).
Feb 12 2019, 12:54 AM
mav committed rS344032: MFC r343562, r343563: Reimplement BIO_ORDERED handling in nvd(4)..
MFC r343562, r343563: Reimplement BIO_ORDERED handling in nvd(4).
Feb 12 2019, 12:53 AM
mav accepted D19148: UMA: unsign some variables related to allocation in hash_alloc()..
Feb 12 2019, 12:43 AM

Feb 11 2019

mav added a comment to D19148: UMA: unsign some variables related to allocation in hash_alloc()..
In D19148#409623, @pfg wrote:

I thought of that but I saw a -1 when calculating it:

hash->uh_hashmask = hash->uh_hashsize - 1;

.A hashsize of zero makes little sense but it is stil a valid value (?).

Feb 11 2019, 4:29 PM
mav added a comment to D19148: UMA: unsign some variables related to allocation in hash_alloc()..

Generally I agree with it. I spotted it too when worked on my change. It just needs 8 times more RAM then what we have now to overflow, so I preferred my patch to be minimal. Just going that way, why not make uh_hashmask also unsigned? Signed bit mask makes even less sense. You've changed hval from int in one place, but not other(s).

Feb 11 2019, 3:44 PM
mav committed rS344009: MFC r343727: Use switch instead of chained if/else to improve readability..
MFC r343727: Use switch instead of chained if/else to improve readability.
Feb 11 2019, 2:50 PM
mav committed rS344008: MFC r343727: Use switch instead of chained if/else to improve readability..
MFC r343727: Use switch instead of chained if/else to improve readability.
Feb 11 2019, 2:50 PM
mav committed rS344007: MFC r343728: Check element type before setting LEDs..
MFC r343728: Check element type before setting LEDs.
Feb 11 2019, 2:49 PM
mav committed rS344006: MFC r343728: Check element type before setting LEDs..
MFC r343728: Check element type before setting LEDs.
Feb 11 2019, 2:48 PM
mav committed rS344005: MFC r343745, r343752: Add missed tunables/sysctls for some new vdev variables..
MFC r343745, r343752: Add missed tunables/sysctls for some new vdev variables.
Feb 11 2019, 2:47 PM
mav committed rS344004: MFC r343745, r343752: Add missed tunables/sysctls for some new vdev variables..
MFC r343745, r343752: Add missed tunables/sysctls for some new vdev variables.
Feb 11 2019, 2:47 PM

Feb 9 2019

mav committed rS343926: MFC r343673: Fix integer math overflow in UMA hash_alloc()..
MFC r343673: Fix integer math overflow in UMA hash_alloc().
Feb 9 2019, 2:10 AM
mav committed rS343925: MFC r343673: Fix integer math overflow in UMA hash_alloc()..
MFC r343673: Fix integer math overflow in UMA hash_alloc().
Feb 9 2019, 2:09 AM

Feb 8 2019

mav added a comment to D18774: Add AMD NTB Hardware Driver for AMD SoC.

Yes, please make a patch to split allocation and programming.

Feb 8 2019, 2:18 PM

Feb 7 2019

mav added a comment to D19109: CID 1009492: Logically dead code in sys/cam/scsi/scsi_xpt.c.

I think I would remove this code completely, together with above NULL assignments. One way or another, this check is at least incomplete, since just in if block before execution can go to out bypassing this check.

Feb 7 2019, 9:11 PM
mav accepted D18785: Fix quotas for UFS after r339008.

It looks like a workaround on workaround, but considering it is deep legacy, so be it.

Feb 7 2019, 8:59 PM

Feb 4 2019

mav committed rS343752: s/Maximal/Maximum/ in sysctl description..
s/Maximal/Maximum/ in sysctl description.
Feb 4 2019, 8:09 PM
mav committed rS343745: Add missed tunables/sysctls for some new vdev variables..
Add missed tunables/sysctls for some new vdev variables.
Feb 4 2019, 4:14 PM
mav added inline comments to D18774: Add AMD NTB Hardware Driver for AMD SoC.
Feb 4 2019, 3:23 PM
mav committed rS343728: Check element type before setting LEDs..
Check element type before setting LEDs.
Feb 4 2019, 1:24 AM
mav closed D19052: usr.sbin/sesutil: Check element type before setting LEDs.
Feb 4 2019, 1:24 AM
mav committed rS343727: Use switch instead of chained if/else to improve readability..
Use switch instead of chained if/else to improve readability.
Feb 4 2019, 1:21 AM
mav closed D19051: cam/scsi: Use switch instead of chained if/else.
Feb 4 2019, 1:21 AM
mav accepted D18924: bridge: Fix spurious warnings about capabilities.
Feb 4 2019, 12:34 AM

Feb 2 2019

mav committed rS343673: Fix integer math overflow in UMA hash_alloc()..
Fix integer math overflow in UMA hash_alloc().
Feb 2 2019, 4:12 AM

Feb 1 2019

mav added inline comments to D18774: Add AMD NTB Hardware Driver for AMD SoC.
Feb 1 2019, 2:45 PM
mav accepted D19052: usr.sbin/sesutil: Check element type before setting LEDs.
Feb 1 2019, 4:14 AM
mav accepted D19051: cam/scsi: Use switch instead of chained if/else.

Looks good to me.

Feb 1 2019, 3:50 AM

Jan 31 2019

mav committed rS343625: Release Sean Eric Fagan (sef) from mentorship..
Release Sean Eric Fagan (sef) from mentorship.
Jan 31 2019, 10:20 PM