smh (Steven Hartland)
User

Projects

User Details

User Since
Aug 3 2014, 6:44 PM (147 w, 1 d)

Recent Activity

Thu, May 18

smh committed rS318438: Revert the partial MFC of r313045 which broke dtrace.
Revert the partial MFC of r313045 which broke dtrace
Thu, May 18, 3:32 AM
smh closed D10769: Revert MFC of r313045 from stable/10 by committing rS318438: Revert the partial MFC of r313045 which broke dtrace.
Thu, May 18, 3:32 AM

Wed, May 17

smh added a reviewer for D10769: Revert MFC of r313045 from stable/10: markj.
Wed, May 17, 8:31 AM
smh created D10769: Revert MFC of r313045 from stable/10.
Wed, May 17, 8:30 AM

Apr 26 2017

smh committed rS317470: MFC r315449:.
MFC r315449:
Apr 26 2017, 10:25 PM
smh committed rS317469: MFC r316460:.
MFC r316460:
Apr 26 2017, 10:23 PM

Apr 24 2017

smh committed rS317375: Partial MFC r316676 and the required r313045.
Partial MFC r316676 and the required r313045
Apr 24 2017, 4:31 PM
smh committed rS317368: MFC r316676:.
MFC r316676:
Apr 24 2017, 11:34 AM

Apr 19 2017

smh accepted D10425: net: Add new 10G and 25G media types to if_media.h.

LGTM

Apr 19 2017, 3:40 AM · network

Apr 14 2017

smh committed rS316944: MFC r316313, r316328:.
MFC r316313, r316328:
Apr 14 2017, 10:02 PM
smh committed rS316943: MFC r303863:.
MFC r303863:
Apr 14 2017, 9:49 PM

Apr 11 2017

smh accepted D10359: loader.efi: only fetch zfs pool guid for the actual boot device.

LGM, much cleaner than it was.

Apr 11 2017, 2:03 PM
smh requested changes to D10359: loader.efi: only fetch zfs pool guid for the actual boot device.
Apr 11 2017, 12:34 PM
smh requested changes to D10293: ixgbe(4): Update HEAD (p2) to 3.2.12-k.

Wow that's a lot of changes.

Apr 11 2017, 8:56 AM

Apr 10 2017

smh committed rS316676: Use estimated RTT for receive buffer auto resizing instead of timestamps.
Use estimated RTT for receive buffer auto resizing instead of timestamps
Apr 10 2017, 8:19 AM
smh closed D9668: Support estimated RTT for receive buffer auto resizing by committing rS316676: Use estimated RTT for receive buffer auto resizing instead of timestamps.
Apr 10 2017, 8:19 AM

Apr 8 2017

smh updated the diff for D9668: Support estimated RTT for receive buffer auto resizing.

Convert timestamp to ticks to ensure that comparisons are correct for
hz != 1000.

Apr 8 2017, 10:28 PM

Apr 7 2017

smh added a comment to D9668: Support estimated RTT for receive buffer auto resizing.

Lawrence is now happy with this in is current form, so just a prod to see if I get any comments for or against from transport members / gnn?

Apr 7 2017, 5:56 PM

Apr 3 2017

smh committed rS316460: Fix expandsz 16.0E vals and vdev_min_asize of RAIDZ children.
Fix expandsz 16.0E vals and vdev_min_asize of RAIDZ children
Apr 3 2017, 1:11 PM

Mar 31 2017

smh updated the diff for D9668: Support estimated RTT for receive buffer auto resizing.

Extract common logic to tcp_autorcvbuf as suggested by lstewart.

Mar 31 2017, 4:27 PM
smh committed rS316328: Allow explicitly assigned IPv6 loopback address to be used in jails.
Allow explicitly assigned IPv6 loopback address to be used in jails
Mar 31 2017, 9:11 AM
smh committed rS316313: Allow explicitly assigned IPv4 loopback address to be used in jails.
Allow explicitly assigned IPv4 loopback address to be used in jails
Mar 31 2017, 12:42 AM

Mar 30 2017

smh accepted D10202: Add a new parameter to restrict jails from binding to privileged ports.
Mar 30 2017, 11:58 PM
smh accepted D10202: Add a new parameter to restrict jails from binding to privileged ports.

Some style nits but otherwise LGM

Mar 30 2017, 3:23 PM

Mar 23 2017

smh committed rS315855: MFC r315423:.
MFC r315423:
Mar 23 2017, 10:43 AM

Mar 19 2017

smh added a comment to D9668: Support estimated RTT for receive buffer auto resizing.

Thanks for the feedback Lawrence, based on that I've updated to use just the SRTT check, added the fastpath version and generally cleaned up.

Mar 19 2017, 12:50 PM
smh updated the diff for D9668: Support estimated RTT for receive buffer auto resizing.

Corrected dtrace autoresize mbuf parameter definition.

Mar 19 2017, 12:43 PM
smh added a comment to D9668: Support estimated RTT for receive buffer auto resizing.

Updated dtrace:

dtrace
#!/usr/sbin/dtrace -s
Mar 19 2017, 12:37 PM
smh updated the diff for D9668: Support estimated RTT for receive buffer auto resizing.

Eliminated timestamp specific code path resulting in a simplified yet still effective single SRTT path.

Mar 19 2017, 12:29 PM

Mar 17 2017

smh closed D10012: Reduce ARC fragmentaion threshold by committing rS315449: Reduce ARC fragmentation threshold.
Mar 17 2017, 12:35 PM
smh committed rS315449: Reduce ARC fragmentation threshold.
Reduce ARC fragmentation threshold
Mar 17 2017, 12:35 PM
smh added a reviewer for D10012: Reduce ARC fragmentaion threshold: avg.
Mar 17 2017, 12:14 PM

Mar 16 2017

smh committed rS315423: Fix libsysdecode vmprot flag decoding.
Fix libsysdecode vmprot flag decoding
Mar 16 2017, 8:55 PM
smh added inline comments to D9757: loader: remove open_disk cache.
Mar 16 2017, 10:27 AM

Mar 15 2017

smh added a reviewer for D10012: Reduce ARC fragmentaion threshold: mav.
Mar 15 2017, 12:40 PM
smh created D10012: Reduce ARC fragmentaion threshold.
Mar 15 2017, 12:40 PM

Mar 14 2017

smh accepted D9999: zfs: add zio_buf_alloc_nowait and use it in vdev_queue_aggregate.

LGTM

Mar 14 2017, 1:46 PM

Mar 7 2017

smh requested changes to D9851: ixgbe(4): Update HEAD to 3.2.12-k.

Massive amounts of changes in this so impossible to see if everything is good, however a number of style related bits highlighted, some of which are regressions.

Mar 7 2017, 5:19 PM
smh accepted D9829: Add a summary line of the ZFS Compressed ARC to top(1).
Mar 7 2017, 4:39 PM

Mar 4 2017

smh accepted D9847: Try to extract the RFC1048 data from PXE.
Mar 4 2017, 3:55 AM

Mar 3 2017

smh requested changes to D9847: Try to extract the RFC1048 data from PXE.
Mar 3 2017, 4:32 PM
smh added a comment to D9847: Try to extract the RFC1048 data from PXE.

book -> bool

Mar 3 2017, 6:56 AM
smh added inline comments to D9847: Try to extract the RFC1048 data from PXE.
Mar 3 2017, 6:53 AM

Mar 2 2017

smh requested changes to D9847: Try to extract the RFC1048 data from PXE.
Mar 2 2017, 10:47 PM
smh requested changes to D9847: Try to extract the RFC1048 data from PXE.
Mar 2 2017, 10:02 PM
smh requested changes to D9847: Try to extract the RFC1048 data from PXE.
Mar 2 2017, 5:15 PM

Feb 28 2017

smh added a comment to D9829: Add a summary line of the ZFS Compressed ARC to top(1).
In D9829#203014, @jhb wrote:

I agree with shortening the line a bit so it fits in 80 cols. If you are worried about duplicating "ARC:" you could perhaps just use a whitespace prefix so you end up with something like:

Mem: 754M Active, 836M Inact, 75M Laundry, 3761M Wired, 10G Free
ARC: 2476M Total, 914M MFU, 1302M MRU, 1696K Anon, 140M Header, 118M Other
     1899M Compressed, 6065M Uncompressed, 2.45:1 Ratio, 318M Overhead
Swap: 3072M Total, 3072M Free
Feb 28 2017, 11:19 PM
smh added a comment to D9829: Add a summary line of the ZFS Compressed ARC to top(1).

Some may find it useful if it was bit shorter, currently its 82 chars in the example.

Feb 28 2017, 4:54 PM

Feb 24 2017

smh added a comment to D9611: ZFS ARC: remove TRIM-ing of cache (L2ARC) devices.

Indeed, do you have any bandwidth to do proper testing to prove either way or have you already done this?

Feb 24 2017, 10:10 AM

Feb 23 2017

smh accepted D9485: Make LACP behavior more closely conform to the standard.

LGTM

Feb 23 2017, 5:39 PM
smh accepted D9757: loader: remove open_disk cache.

LGTM

Feb 23 2017, 9:13 AM

Feb 22 2017

smh accepted D9706: loader: update symlink support in zfs reader.

LGTM

Feb 22 2017, 1:42 PM
smh added inline comments to D9706: loader: update symlink support in zfs reader.
Feb 22 2017, 12:32 PM
smh requested changes to D9706: loader: update symlink support in zfs reader.
Feb 22 2017, 9:31 AM
smh added a reviewer for D9668: Support estimated RTT for receive buffer auto resizing: hiren.
Feb 22 2017, 9:17 AM

Feb 21 2017

smh accepted D7421: reimplement zfsctl (.zfs) support.
Feb 21 2017, 5:30 PM
smh requested changes to D7421: reimplement zfsctl (.zfs) support.

Just a few seemingly redundant assignments to error vars, sorry didn't spot them before

Feb 21 2017, 11:24 AM

Feb 20 2017

smh added inline comments to D9668: Support estimated RTT for receive buffer auto resizing.
Feb 20 2017, 9:14 AM
smh updated the diff for D9668: Support estimated RTT for receive buffer auto resizing.

Disabled reset receive buffer auto scaling when not in bulk receive mode, which gives an extra 20% performance increase, bring it closer to Linux.

Feb 20 2017, 9:08 AM

Feb 19 2017

smh updated D9668: Support estimated RTT for receive buffer auto resizing.
Feb 19 2017, 1:43 AM
smh added a comment to D9668: Support estimated RTT for receive buffer auto resizing.

For reference fastpath version of this hasn't been done, which need also need the same changes to:

tcp_stacks/fastpath.c
Feb 19 2017, 1:39 AM
smh updated D9668: Support estimated RTT for receive buffer auto resizing.
Feb 19 2017, 1:22 AM
smh retitled D9668: Support estimated RTT for receive buffer auto resizing from to Support estimated RTT for receive buffer auto resizing.
Feb 19 2017, 1:21 AM

Feb 15 2017

smh added a comment to D9611: ZFS ARC: remove TRIM-ing of cache (L2ARC) devices.
In D9611#198637, @avg wrote:

Essentially, L2ARC periodically writes to every block on the device and that allows to physically move the data around.
That's exactly the reason why I think that TRIM is not needed for L2ARC.
TRIM is useful when we don't need data in some area, but we are not going to overwrite that area, so we need to a way to tell the storage system that it can reuse the physical cells without worrying about any data in them. But if we overwrite that area anyway, then the storage system is automatically aware that the data in those physical cells is obsolete. It's free to choose either those same cells or any different cells for the new data according to the wear leveling algorithms, but that's beside the point.

Feb 15 2017, 5:18 PM
smh added a comment to D9611: ZFS ARC: remove TRIM-ing of cache (L2ARC) devices.
In D9611#198616, @avg wrote:

My understanding of how L2ARC writing works is this. The code maintains a "hand" (like a clock hand) that points to a disk offset. At the regular intervals a certain space in front of the hand is freed by discarding L2 headers that point to that space and then new buffers are written to that space. Then the hand is moved forward by the appropriate amount.
There is also some freeing of L2 headers when ARC headers are freed, etc. In any case, after some uptime almost the whole cache disk is usually filled with data. And the hand inevitably moves forward. So, every block gets written over sooner or later. I do not see how TRIM helps in that case.
The only scenario where it makes difference, in my humble opinion, is the scenario that @mav described: a worn out disk where the "cheese holes" behind the hand can make some difference for writing new blocks at the hand. But I think that that's too marginal to be important.

Feb 15 2017, 3:55 PM
smh added a comment to D9611: ZFS ARC: remove TRIM-ing of cache (L2ARC) devices.
In D9611#198582, @mav wrote:
In D9611#198579, @smh wrote:

In scenario #1 the performance of TRIM is also generally good, mitigating the need to avoid doing it.

Is there such thing as good TRIM performance? On my new Samsung 950 NVMe I had to disable TRIM as unusable. Though yes, NVMe driver still probably needs aggregation of TRIM requests to get better numbers.

Feb 15 2017, 3:11 PM
smh added a comment to D9611: ZFS ARC: remove TRIM-ing of cache (L2ARC) devices.

In scenario #1 the performance of TRIM is also generally good, mitigating the need to avoid doing it.

Feb 15 2017, 2:45 PM
smh added a comment to D9611: ZFS ARC: remove TRIM-ing of cache (L2ARC) devices.

This can could excessive slow down as the capacity of the disk is reached, however it could be argued that a better mitigation for L2ARC devices would be to use an under-provisioned slice to ensure the SSD controller always has space to work.

Feb 15 2017, 2:23 PM

Jan 23 2017

smh requested changes to D9165: loader.efi environment related cleanups.
Jan 23 2017, 9:51 AM

Jan 16 2017

smh committed rS312279: MFC r311769:.
MFC r311769:
Jan 16 2017, 9:16 AM
smh committed rS312278: MFC r311769:.
MFC r311769:
Jan 16 2017, 9:13 AM

Jan 11 2017

smh accepted D9144: Use proper prototypes in struct boot_module_t.
Jan 11 2017, 7:50 PM

Jan 9 2017

smh committed rS311769: Fix rstat: symbol not in namelist from netstat.
Fix rstat: symbol not in namelist from netstat
Jan 9 2017, 9:28 AM

Nov 28 2016

smh accepted D7421: reimplement zfsctl (.zfs) support.
Nov 28 2016, 11:41 AM
smh added a comment to D7421: reimplement zfsctl (.zfs) support.

Couple of little style nits and I'd like to understand why ENAMETOOLONG error gets turned into success in a few places.

Nov 28 2016, 9:58 AM

Nov 2 2016

smh accepted D8403: mpr/mps device mapper override.
Nov 2 2016, 9:28 AM

Oct 31 2016

smh requested changes to D8403: mpr/mps device mapper override.

Wouldn't it be better to have 0 (the default) be "use the PhyNum field as a fallback to the mapping logic" as that way at least the device would initialise when mpX_mapping_get_sas_id fails?

Oct 31 2016, 11:10 PM

Aug 15 2016

smh committed rS304112: MFS r304067:.
MFS r304067:
Aug 15 2016, 8:07 AM

Aug 14 2016

smh committed rS304081: MFC r303971:.
MFC r303971:
Aug 14 2016, 3:28 PM
smh committed rS304067: MFC r303971:.
MFC r303971:
Aug 14 2016, 12:40 AM

Aug 11 2016

smh committed rS303971: Fix vtnet hang with max_virtqueue_pairs > VTNET_MAX_QUEUE_PAIRS.
Fix vtnet hang with max_virtqueue_pairs > VTNET_MAX_QUEUE_PAIRS
Aug 11 2016, 9:14 PM

Aug 4 2016

smh added inline comments to D6533: zfs: step #1 of bringing zpl to full obedience of vfs.
Aug 4 2016, 9:32 PM
smh added inline comments to D6533: zfs: step #1 of bringing zpl to full obedience of vfs.
Aug 4 2016, 9:09 PM

Jul 22 2016

smh added a comment to D6533: zfs: step #1 of bringing zpl to full obedience of vfs.

Thanks for attacking this Andriy its a big task which really needed some attention.

Jul 22 2016, 2:31 PM

Jul 13 2016

smh committed rS302714: MFC r302265, r302382.
MFC r302265, r302382
Jul 13 2016, 8:34 AM
smh added a comment to D7201: Retry reading from boot disk if first access fails (HP BIOS bug??).

Apart from the comment this looks good in principle, however I think we need to better understand why retrying the probe works as it feels like where just hiding the real error with this.

Jul 13 2016, 8:19 AM

Jul 11 2016

smh added a comment to D7179: move handling of zvol devices out of txg sync thread.

This looks reasonable however as you say its potentially racey with multiple renames happening.

Jul 11 2016, 1:02 PM

Jul 6 2016

smh committed rS302382: Fix ZFS ARC min / max tunable.
Fix ZFS ARC min / max tunable
Jul 6 2016, 11:49 PM

Jun 29 2016

smh committed rS302265: Allow ZFS ARC min / max to be tuned at runtime.
Allow ZFS ARC min / max to be tuned at runtime
Jun 29 2016, 7:56 AM
smh closed D5907: Allow ZFS ARC min / max to be tuned at runtime by committing rS302265: Allow ZFS ARC min / max to be tuned at runtime.
Jun 29 2016, 7:56 AM

Jun 28 2016

smh added a reviewer for D5907: Allow ZFS ARC min / max to be tuned at runtime: allanjude.
Jun 28 2016, 3:14 PM
smh added inline comments to D5907: Allow ZFS ARC min / max to be tuned at runtime.
Jun 28 2016, 3:14 PM

Jun 14 2016

smh added a comment to D6780: loader.efi try to fallback on first device.

I'd like to know why finding the current device would ever fail?

Jun 14 2016, 2:50 AM

Jun 3 2016

smh accepted D6709: Use the UEFI timer on arm64,the RTC may not be available.
Jun 3 2016, 9:42 PM

Jun 1 2016

smh committed rS301131: Fix tzsetup not installing /etc/localtime for UTC.
Fix tzsetup not installing /etc/localtime for UTC
Jun 1 2016, 3:39 PM

May 9 2016

smh added inline comments to D6153: Speed up vdev_geom_open_by_guids.
May 9 2016, 4:50 PM
smh added inline comments to D6153: Speed up vdev_geom_open_by_guids.
May 9 2016, 9:31 AM

May 5 2016

smh added inline comments to D6153: Speed up vdev_geom_open_by_guids.
May 5 2016, 4:03 PM

Apr 11 2016

smh committed rS297819: Only include sysctl in kernel build.
Only include sysctl in kernel build
Apr 11 2016, 1:17 PM
smh updated the diff for D5907: Allow ZFS ARC min / max to be tuned at runtime.

Remove redundent check on <= 0

Apr 11 2016, 9:12 AM