Page MenuHomeFreeBSD

mav (Alexander Motin)
User

Projects

User Details

User Since
Jun 3 2014, 6:27 PM (299 w, 2 d)

Recent Activity

Today

mav added a comment to D23852: scsi_scan_bus: ignore Initiator ID for SAS transport in.

It is not a problem of UINT_MAX, it is a problem to ignore potentially correct reported value.

Thu, Feb 27, 8:48 PM
mav added a comment to D23852: scsi_scan_bus: ignore Initiator ID for SAS transport in.

I agree with everything you say we should do. But we are not there yet, and until that time SIMs have to keep some sort of mapping table for CAM to use. From the CAM point now all target IDs of the SCSI domain the HBA is connected to, including the initiator_id are from that mapping table. I am not sure what exactly makes no sense to you in my descrintion of how things are working for FC now and why exactly the same is not applicable to SAS to add this dirty hack?

Thu, Feb 27, 7:51 PM
mav accepted D23854: Unbreak openzfs ports.

I have no objections, but it would be good to request review from some port committers.

Thu, Feb 27, 7:28 PM
mav added a comment to D23852: scsi_scan_bus: ignore Initiator ID for SAS transport in.

The patch will make XPT_SCAN_TGT/XPT_SCAN_BUS to scan target they should not.

Thu, Feb 27, 7:17 PM
mav added a comment to D23852: scsi_scan_bus: ignore Initiator ID for SAS transport in.

As I have told, SAS may also work as target, we just miss respective SIM. And the code skipping the initiaror_id during scan allows to not scan own target, if we ever have non-FC targets. What is so difficult in fixing few SAS SIMs to report initiator_id out of the range of valid initiator role targets?

Thu, Feb 27, 6:19 PM
mav requested changes to D23852: scsi_scan_bus: ignore Initiator ID for SAS transport in.

I think it is a wrong way to go. While specific initiator ID value has sense only on parallel SCSI and loop mode FC and is obsolete these days, the idea of having some id reserved for the local HBA itself makes a lot of sense for FC now, when isp(4) driver can operate both initiator and target role same time, and the initiator id is used to represent target role periphs, while other ids are used rot initiator role. Such concepts looks totally valid to me. Even though we do not have SAS target driver (I dreamed about it for years), I don't think we should make it impossible instead of trivially fixing few drivers.

Thu, Feb 27, 4:09 PM

Yesterday

mav committed rS358357: MFZoL: Relax restriction on zfs_ioc_next_obj() iteration.
MFZoL: Relax restriction on zfs_ioc_next_obj() iteration
Wed, Feb 26, 8:39 PM
mav committed rS358342: MFZoL: Fix resilver writes in vdev_indirect_io_start.
MFZoL: Fix resilver writes in vdev_indirect_io_start
Wed, Feb 26, 4:52 PM
mav committed rS358340: Fix patch mismerge in r358336..
Fix patch mismerge in r358336.
Wed, Feb 26, 4:04 PM
mav committed rS358339: MFZoL: Fix issue with scanning dedup blocks as scan ends.
MFZoL: Fix issue with scanning dedup blocks as scan ends
Wed, Feb 26, 4:00 PM
mav committed rS358337: MFZoL: Fix 2 small bugs with cached dsl_scan_phys_t.
MFZoL: Fix 2 small bugs with cached dsl_scan_phys_t
Wed, Feb 26, 3:48 PM
mav committed rS358336: MFZoL: Fix txg_sync_thread hang in scan_exec_io().
MFZoL: Fix txg_sync_thread hang in scan_exec_io()
Wed, Feb 26, 3:45 PM

Tue, Feb 25

mav committed rS358313: MFC r349381: Avoid extra taskq_dispatch() calls by DMU..
MFC r349381: Avoid extra taskq_dispatch() calls by DMU.
Tue, Feb 25, 3:53 PM
mav committed rS358312: MFC r349381: Avoid extra taskq_dispatch() calls by DMU..
MFC r349381: Avoid extra taskq_dispatch() calls by DMU.
Tue, Feb 25, 3:53 PM

Sat, Feb 22

mav accepted D23792: Add 'ctld -t'..
Sat, Feb 22, 6:53 PM
mav committed rS358238: MFC r357685: Bind CTL backends taskqueues to the CTL process..
MFC r357685: Bind CTL backends taskqueues to the CTL process.
Sat, Feb 22, 4:49 AM

Fri, Feb 21

mav committed rS358204: MFC r357657: Remove duplicate dbufs accounting..
MFC r357657: Remove duplicate dbufs accounting.
Fri, Feb 21, 4:33 AM
mav committed rS358201: MFC r357920: Add support for Hygon NTB PCI device in ntb_hw_amd driver..
MFC r357920: Add support for Hygon NTB PCI device in ntb_hw_amd driver.
Fri, Feb 21, 4:30 AM
mav committed rS358200: MFC r357919: Add Hygon PCI ID and description for AHCI SATA controller..
MFC r357919: Add Hygon PCI ID and description for AHCI SATA controller.
Fri, Feb 21, 4:29 AM
mav committed rS358199: MFC r357919: Add Hygon PCI ID and description for AHCI SATA controller..
MFC r357919: Add Hygon PCI ID and description for AHCI SATA controller.
Fri, Feb 21, 4:29 AM

Thu, Feb 20

mav committed rS358137: MFC r357639: Reduce number of atomic_add() calls in aggsum..
MFC r357639: Reduce number of atomic_add() calls in aggsum.
Thu, Feb 20, 1:04 AM

Tue, Feb 18

mav committed rS358060: MFC r357502: Few microoptimizations to dbuf layer..
MFC r357502: Few microoptimizations to dbuf layer.
Tue, Feb 18, 1:22 AM
mav committed rS358059: MFC r357453: Unblock kstat.zfs.misc.dbufstats sysctls..
MFC r357453: Unblock kstat.zfs.misc.dbufstats sysctls.
Tue, Feb 18, 1:19 AM

Fri, Feb 14

mav committed rS357920: Add support for Hygon NTB PCI device in ntb_hw_amd driver..
Add support for Hygon NTB PCI device in ntb_hw_amd driver.
Fri, Feb 14, 3:05 PM
mav closed D23565: ntb: Add Hygon Dhyana support.
Fri, Feb 14, 3:05 PM
mav accepted D23565: ntb: Add Hygon Dhyana support.
Fri, Feb 14, 3:00 PM
mav added inline comments to D23565: ntb: Add Hygon Dhyana support.
Fri, Feb 14, 2:58 PM
mav committed rS357919: Add Hygon PCI ID and description for AHCI SATA controller..
Add Hygon PCI ID and description for AHCI SATA controller.
Fri, Feb 14, 2:55 PM
mav closed D23556: ahci: Add Hygon Dhyana support.
Fri, Feb 14, 2:55 PM
mav accepted D23556: ahci: Add Hygon Dhyana support.
Fri, Feb 14, 2:55 PM

Tue, Feb 11

mav committed rS357747: Enable commit access for Ryan Moeller (freqlabs)..
Enable commit access for Ryan Moeller (freqlabs).
Tue, Feb 11, 12:44 AM

Sat, Feb 8

mav committed rS357685: Bind CTL backends taskqueues to the CTL process..
Bind CTL backends taskqueues to the CTL process.
Sat, Feb 8, 10:00 PM
mav committed rS357684: MFC r353678 (by avg):.
MFC r353678 (by avg):
Sat, Feb 8, 9:56 PM

Fri, Feb 7

mav committed rS357657: Remove duplicate dbufs accounting..
Remove duplicate dbufs accounting.
Fri, Feb 7, 3:51 PM
mav committed rS357656: MFC r326956 (by cem):.
MFC r326956 (by cem):
Fri, Feb 7, 3:28 PM
mav added inline comments to D23565: ntb: Add Hygon Dhyana support.
Fri, Feb 7, 2:38 PM
mav added a comment to D23556: ahci: Add Hygon Dhyana support.

Hi. Does the Hygon really have only one AHCI ID where AMD had 5?

Fri, Feb 7, 2:33 PM

Thu, Feb 6

mav committed rS357639: Reduce number of atomic_add() calls in aggsum..
Reduce number of atomic_add() calls in aggsum.
Thu, Feb 6, 8:33 PM
mav added a comment to D23541: Make BIO_SPEEDUP opt-in.

I won't object too much, but first addition of quite specialized command and then workaround for it does not make me particularly happy.

Thu, Feb 6, 8:27 PM
mav added a comment to D23541: Make BIO_SPEEDUP opt-in.

I would honestly prefer drivers to report reasonable errors on unknown commands. That would fix the issue once and for all.

Thu, Feb 6, 8:19 PM
mav accepted D23483: Fix for unbalanced EPOCH(9) usage in kernel interrupt handler.
Thu, Feb 6, 4:23 PM

Wed, Feb 5

mav committed rS357555: MFC r356993: Update route MTUs for bridge, lagg and vlan interfaces..
MFC r356993: Update route MTUs for bridge, lagg and vlan interfaces.
Wed, Feb 5, 12:47 AM

Tue, Feb 4

mav committed rS357502: Few microoptimizations to dbuf layer..
Few microoptimizations to dbuf layer.
Tue, Feb 4, 3:54 PM
mav accepted D23483: Fix for unbalanced EPOCH(9) usage in kernel interrupt handler.

Looks good to me.

Tue, Feb 4, 3:43 PM
mav accepted D23483: Fix for unbalanced EPOCH(9) usage in kernel interrupt handler.
Tue, Feb 4, 3:23 PM
mav added a comment to D23483: Fix for unbalanced EPOCH(9) usage in kernel interrupt handler.

I am not sure what is the point of atomic there, except may be preventing compiler from optimizing it out, if that can happen somehow. Otherwise it looks fine to me. I don't remember the interrupt teardown semantics here to say whether use-after-free is possible, but I agree with @galatin that flags may change in run time, causing epoch leak, so this code must be here.

Tue, Feb 4, 3:23 PM

Mon, Feb 3

mav added reviewers for D23478: rework how ZVOLs are updated in response to DSL operations: mmacy, ryan_freqlabs.com.
Mon, Feb 3, 6:31 PM
mav committed rS357453: Unblock kstat.zfs.misc.dbufstats sysctls..
Unblock kstat.zfs.misc.dbufstats sysctls.
Mon, Feb 3, 5:10 PM

Thu, Jan 30

mav committed rS357297: MFC r356707: Map ECKSUM and EFRAGS from ZFS onto real errnos..
MFC r356707: Map ECKSUM and EFRAGS from ZFS onto real errnos.
Thu, Jan 30, 4:39 PM
mav committed rS357296: MFC r343111,343113-343115 (by mckusick):.
MFC r343111,343113-343115 (by mckusick):
Thu, Jan 30, 4:11 PM

Tue, Jan 28

mav accepted D22077: Speed up "zpool import" in the presence of many zvols.

I have no objections, other than I would not use braces for the if's, or at least put the opening one in the first case on the same like, as Ryan proposed.

Tue, Jan 28, 10:31 PM

Jan 22 2020

mav committed rS356993: Update route MTUs for bridge, lagg and vlan interfaces..
Update route MTUs for bridge, lagg and vlan interfaces.
Jan 22 2020, 8:37 PM
mav committed rS356961: MFC r356474, r356480, r356482, r356506:.
MFC r356474, r356480, r356482, r356506:
Jan 22 2020, 1:08 AM
mav committed rS356960: MFC r351357 (by imp): Document RST support in nvme(4) and ahci(4)..
MFC r351357 (by imp): Document RST support in nvme(4) and ahci(4).
Jan 22 2020, 1:04 AM
mav committed rS356959: MFC r347967 (by cem): nvd.4: Reference nda(4).
MFC r347967 (by cem): nvd.4: Reference nda(4)
Jan 22 2020, 1:03 AM
mav committed rS356958: MFC r341709 (by imp): Add nda(4) cross reference to nvme(4).
MFC r341709 (by imp): Add nda(4) cross reference to nvme(4)
Jan 22 2020, 12:52 AM
mav committed rS356957: MFC r355774 (by mmel): Properly synchronize completion DMA buffers..
MFC r355774 (by mmel): Properly synchronize completion DMA buffers.
Jan 22 2020, 12:49 AM
mav committed rS356956: MFC r355721 (by imp): Move to using bool instead of boolean_t.
MFC r355721 (by imp): Move to using bool instead of boolean_t
Jan 22 2020, 12:46 AM
mav committed rS356955: MFC r355631 (by imp): Move reset to the interrutp processing stage.
MFC r355631 (by imp): Move reset to the interrutp processing stage
Jan 22 2020, 12:44 AM
mav committed rS356954: MFC r355465 (by imp): trackers always know what qpair they are on.
MFC r355465 (by imp): trackers always know what qpair they are on
Jan 22 2020, 12:41 AM

Jan 17 2020

mav committed rS356825: MFC r355884: Add missing "ereport." prefixes of ZFS events..
MFC r355884: Add missing "ereport." prefixes of ZFS events.
Jan 17 2020, 4:13 AM
mav committed rS356824: MFC r356315: Avoid duplicate I/O statistics accounting..
MFC r356315: Avoid duplicate I/O statistics accounting.
Jan 17 2020, 4:12 AM
mav committed rS356823: MFC r356315: Avoid duplicate I/O statistics accounting..
MFC r356315: Avoid duplicate I/O statistics accounting.
Jan 17 2020, 4:11 AM

Jan 14 2020

mav committed rS356720: Restore loop break in vm_pageout_lowmem()..
Restore loop break in vm_pageout_lowmem().
Jan 14 2020, 3:28 AM
mav committed rS356712: MFC r356240: Relax locking of carp_forus()..
MFC r356240: Relax locking of carp_forus().
Jan 14 2020, 12:52 AM
mav committed rS356711: MFC r356240: Relax locking of carp_forus()..
MFC r356240: Relax locking of carp_forus().
Jan 14 2020, 12:50 AM

Jan 13 2020

mav committed rS356707: Map ECKSUM and EFRAGS from ZFS onto real errnos..
Map ECKSUM and EFRAGS from ZFS onto real errnos.
Jan 13 2020, 10:06 PM
mav closed D22973: Map ECKSUM and EFRAGS from ZFS onto real errnos.
Jan 13 2020, 10:06 PM

Jan 10 2020

mav committed rS356589: MFC r356214: Avoid few memory accesses in g_disk_done()..
MFC r356214: Avoid few memory accesses in g_disk_done().
Jan 10 2020, 12:55 AM
mav committed rS356588: MFC r356214: Avoid few memory accesses in g_disk_done()..
MFC r356214: Avoid few memory accesses in g_disk_done().
Jan 10 2020, 12:53 AM
mav committed rS356587: MFC r356181: Fix GEOM_VIRSTOR orphanization..
MFC r356181: Fix GEOM_VIRSTOR orphanization.
Jan 10 2020, 12:48 AM
mav committed rS356586: MFC r356181: Fix GEOM_VIRSTOR orphanization..
MFC r356181: Fix GEOM_VIRSTOR orphanization.
Jan 10 2020, 12:47 AM
mav committed rS356585: MFC r356178: Fix GEOM_MOUNTVER orphanization..
MFC r356178: Fix GEOM_MOUNTVER orphanization.
Jan 10 2020, 12:47 AM
mav committed rS356584: MFC r356178: Fix GEOM_MOUNTVER orphanization..
MFC r356178: Fix GEOM_MOUNTVER orphanization.
Jan 10 2020, 12:46 AM
mav committed rS356583: MFC r356162, r356182: Fix GEOM_SHSEC orphanization..
MFC r356162, r356182: Fix GEOM_SHSEC orphanization.
Jan 10 2020, 12:46 AM
mav committed rS356582: MFC r356162, r356182: Fix GEOM_SHSEC orphanization..
MFC r356162, r356182: Fix GEOM_SHSEC orphanization.
Jan 10 2020, 12:45 AM
mav committed rS356581: MFC r356151: Fix GEOM_GATE orphanization..
MFC r356151: Fix GEOM_GATE orphanization.
Jan 10 2020, 12:44 AM
mav committed rS356580: MFC r356151: Fix GEOM_GATE orphanization..
MFC r356151: Fix GEOM_GATE orphanization.
Jan 10 2020, 12:43 AM
mav committed rS356579: MFC r356138: Fix GEOM_UZIP orphanization..
MFC r356138: Fix GEOM_UZIP orphanization.
Jan 10 2020, 12:43 AM
mav committed rS356578: MFC r356138: Fix GEOM_UZIP orphanization..
MFC r356138: Fix GEOM_UZIP orphanization.
Jan 10 2020, 12:42 AM
mav committed rS356577: MFC r356138: Reimplement gvinum orphanization..
MFC r356138: Reimplement gvinum orphanization.
Jan 10 2020, 12:42 AM
mav committed rS356576: MFC r356108:.
MFC r356108:
Jan 10 2020, 12:41 AM

Jan 9 2020

mav committed rS356531: MFC r356284: Remove extra check for provider being closed..
MFC r356284: Remove extra check for provider being closed.
Jan 9 2020, 12:39 AM
mav committed rS356530: MFC r351824, r356392: Report the Host Buffer Memory minimum and preferred sizes..
MFC r351824, r356392: Report the Host Buffer Memory minimum and preferred sizes.
Jan 9 2020, 12:36 AM

Jan 8 2020

mav committed rS356506: Fix copy-paste bug in HMB free code..
Fix copy-paste bug in HMB free code.
Jan 8 2020, 6:26 PM

Jan 7 2020

mav committed rS356482: Minor adjustments to r356474 and r356480..
Minor adjustments to r356474 and r356480.
Jan 7 2020, 11:30 PM
mav committed rS356480: Increate HMB limit from 1% to 5%..
Increate HMB limit from 1% to 5%.
Jan 7 2020, 11:10 PM
mav added a comment to D22973: Map ECKSUM and EFRAGS from ZFS onto real errnos.

EFRAGS seems not to be really used on FreeBSD. It is mentioned once in code, but it seems to be dead from the day one -- used on Illumos to support dumping to ZVOL. I am not sure why ZoL haven't dropped it yet.
About ECKSUM->EINTEGRITY I have no objections, other then Linux maps it into some nonsense, so some applications may get surprised if notice the difference. But that may be unavoidable.

Jan 7 2020, 10:13 PM
mav committed rS356474: Add Host Memory Buffer support to nvme(4)..
Add Host Memory Buffer support to nvme(4).
Jan 7 2020, 9:17 PM
mav committed rS356430: MFC r356216: Don't spin on cleanup_lock if we are not interrupt..
MFC r356216: Don't spin on cleanup_lock if we are not interrupt.
Jan 7 2020, 12:55 AM
mav committed rS356429: MFC r356216: Don't spin on cleanup_lock if we are not interrupt..
MFC r356216: Don't spin on cleanup_lock if we are not interrupt.
Jan 7 2020, 12:54 AM

Jan 6 2020

mav committed rS356392: Fix host memory buffer sizes reporting..
Fix host memory buffer sizes reporting.
Jan 6 2020, 1:51 AM
mav committed rS356391: MFC r356042: Make pass(4) handle misaligned buffers of MAXPHYS size..
MFC r356042: Make pass(4) handle misaligned buffers of MAXPHYS size.
Jan 6 2020, 1:15 AM
mav committed rS356390: MFC r356042: Make pass(4) handle misaligned buffers of MAXPHYS size..
MFC r356042: Make pass(4) handle misaligned buffers of MAXPHYS size.
Jan 6 2020, 1:12 AM

Jan 3 2020

mav committed rS356315: Avoid duplicate I/O statistics accounting..
Avoid duplicate I/O statistics accounting.
Jan 3 2020, 4:37 AM

Jan 2 2020

mav committed rS356286: MFC r349586 (by rlibby): g_eli_create: only dec g_access acw if we inc'd it.
MFC r349586 (by rlibby): g_eli_create: only dec g_access acw if we inc'd it
Jan 2 2020, 8:35 PM
mav committed rS356285: MFC r349586 (by rlibby): g_eli_create: only dec g_access acw if we inc'd it.
MFC r349586 (by rlibby): g_eli_create: only dec g_access acw if we inc'd it
Jan 2 2020, 8:34 PM
mav committed rS356284: Remove extra check for provider being closed..
Remove extra check for provider being closed.
Jan 2 2020, 8:31 PM
mav committed rS356275: MFC r352658,352677,352713,352728: Fix/improve interrupt threads scheduling..
MFC r352658,352677,352713,352728: Fix/improve interrupt threads scheduling.
Jan 2 2020, 5:27 PM

Dec 31 2019

mav committed rS356240: Relax locking of carp_forus()..
Relax locking of carp_forus().
Dec 31 2019, 6:58 PM