Page MenuHomeFreeBSD

imp (Warner Losh)
User

Projects

User Details

User Since
Jun 2 2014, 4:20 PM (235 w, 6 d)

Recent Activity

Yesterday

imp committed rS341770: Fix typo in powerpcspe name..
Fix typo in powerpcspe name.
Sun, Dec 9, 9:55 PM
imp committed rS341769: Send a START UNIT command when a disk responds with an ASC of 04/1C..
Send a START UNIT command when a disk responds with an ASC of 04/1C.
Sun, Dec 9, 9:38 PM
imp added inline comments to D18481: rc.subr: Implement list_vars without using 'read'.
Sun, Dec 9, 8:28 PM

Sat, Dec 8

imp added inline comments to D18481: rc.subr: Implement list_vars without using 'read'.
Sat, Dec 8, 11:17 PM

Fri, Dec 7

imp added inline comments to D17947: Rework UEFI ESP generation.
Fri, Dec 7, 10:28 PM
imp committed rS341710: Even though they are reserved, cdw2 and cdw3 can be set via nvme-cli.
Even though they are reserved, cdw2 and cdw3 can be set via nvme-cli
Fri, Dec 7, 9:58 PM
imp committed rS341709: Add nda(4) cross reference to nvme(4).
Add nda(4) cross reference to nvme(4)
Fri, Dec 7, 9:57 PM

Thu, Dec 6

imp committed rS341664: Update paths based on last-minute changes from libexec to lib..
Update paths based on last-minute changes from libexec to lib.
Thu, Dec 6, 11:41 PM
imp committed rS341663: Declare global function print_intel_add_smart in header.
Declare global function print_intel_add_smart in header
Thu, Dec 6, 11:29 PM
imp committed rS341662: Use proper prototypes..
Use proper prototypes.
Thu, Dec 6, 11:29 PM
imp committed rS341661: It's useful to have this be a global function..
It's useful to have this be a global function.
Thu, Dec 6, 10:59 PM
imp committed rS341659: Move intel and wdc files to their own modules.
Move intel and wdc files to their own modules
Thu, Dec 6, 10:59 PM
imp committed rS341660: This is not a samsung standard, so remove that alias..
This is not a samsung standard, so remove that alias.
Thu, Dec 6, 10:59 PM
imp closed D18460: Move intel and wdc files to their own modules.
Thu, Dec 6, 10:59 PM
imp committed rS341658: Const poison the command interface.
Const poison the command interface
Thu, Dec 6, 10:59 PM
imp closed D18459: Const poison the command interface.
Thu, Dec 6, 10:59 PM
imp committed rS341657: Dynamically load .so modules to expand functionality.
Dynamically load .so modules to expand functionality
Thu, Dec 6, 10:58 PM
imp closed D18455: Dynamically load .so modules to expand functionality.
Thu, Dec 6, 10:58 PM
imp added a comment to D18455: Dynamically load .so modules to expand functionality.
In D18455#393317, @cem wrote:

LGTM. I still find the 'mirror' name confusing, but maybe that's a personal problem ;-).

Thu, Dec 6, 10:51 PM
imp updated the diff for D18455: Dynamically load .so modules to expand functionality.

updates per cem@

Thu, Dec 6, 9:27 PM
imp added inline comments to D18455: Dynamically load .so modules to expand functionality.
Thu, Dec 6, 9:21 PM
imp added inline comments to D18452: Add device_delete_children(9) manual page.
Thu, Dec 6, 7:25 PM
imp added a reviewer for D18460: Move intel and wdc files to their own modules: cem.
Thu, Dec 6, 6:28 PM
imp created D18460: Move intel and wdc files to their own modules.
Thu, Dec 6, 6:27 PM
imp updated the test plan for D18459: Const poison the command interface.
Thu, Dec 6, 6:26 PM
imp created D18459: Const poison the command interface.
Thu, Dec 6, 6:25 PM
imp updated the test plan for D18455: Dynamically load .so modules to expand functionality.
Thu, Dec 6, 5:19 PM
imp created D18455: Dynamically load .so modules to expand functionality.
Thu, Dec 6, 5:15 PM

Wed, Dec 5

imp added inline comments to D15637: Regularize the declaration of the exit syscall..
Wed, Dec 5, 6:47 PM

Tue, Dec 4

imp updated the diff for D18351: Add BIO_SPEEDUP.

Update the pacing base that Kirk sent me to cope better with multiple workers.

Tue, Dec 4, 10:26 PM
imp accepted D18430: Reduce mountd DNS queries.

lgtm

Tue, Dec 4, 10:22 PM

Mon, Dec 3

imp added inline comments to D17947: Rework UEFI ESP generation.
Mon, Dec 3, 11:46 PM
imp added a comment to D18369: stand/i386: rename .s to .S to use Clang IAS.

I looked at this on my phone a few days ago and meant to stamp it with LGTM when I got back to my computer, but that fell through the cracks.

Mon, Dec 3, 7:22 PM
imp added inline comments to D18408: PR209468: aacraid fixes to work with newer controller firmware.
Mon, Dec 3, 7:15 PM
imp committed rS341433: Move inclusion of src.opts.mk later..
Move inclusion of src.opts.mk later.
Mon, Dec 3, 5:52 PM
imp accepted D18408: PR209468: aacraid fixes to work with newer controller firmware.

The code seems fine, other than the weird msi_tupelo which randomly seems to be the same as msi_enabled, but only some of the time. A comment is needed to explain why the oddity.

Mon, Dec 3, 3:35 PM
imp accepted D17941: acpica : move SRAT/SLIT parsing to sys/dev/acpica.
Mon, Dec 3, 3:26 PM
imp accepted D18414: mkisoimages.sh: don't use -p flag for cp.
Mon, Dec 3, 2:58 PM

Sun, Dec 2

imp committed rS341416: Fix typo in comment.
Fix typo in comment
Sun, Dec 2, 11:14 PM
imp committed rS341415: Delete the undocumented alias 'wds'..
Delete the undocumented alias 'wds'.
Sun, Dec 2, 11:14 PM
imp committed rS341414: Move Intel specific log pages to intel.c.
Move Intel specific log pages to intel.c
Sun, Dec 2, 11:14 PM
imp committed rS341413: Usage cleanup pt 2.
Usage cleanup pt 2
Sun, Dec 2, 11:14 PM
imp committed rS341412: Usage cleanup pt 1.
Usage cleanup pt 1
Sun, Dec 2, 11:14 PM
imp committed rS341411: Return after we find the dispatched function..
Return after we find the dispatched function.
Sun, Dec 2, 11:14 PM
imp committed rS341410: Move the hgst/wdc log page printing code into wdc.c.
Move the hgst/wdc log page printing code into wdc.c
Sun, Dec 2, 11:14 PM
imp committed rS341409: Move common logpage routines into nvmecontrol.h.
Move common logpage routines into nvmecontrol.h
Sun, Dec 2, 11:14 PM
imp committed rS341408: Make logpage functions a linker set..
Make logpage functions a linker set.
Sun, Dec 2, 11:14 PM
imp committed rS341407: Move nvmecontrol to using linker sets for commands.
Move nvmecontrol to using linker sets for commands
Sun, Dec 2, 11:11 PM
imp closed D18403: Move nvmecontrol to using linker sets.
Sun, Dec 2, 11:11 PM
imp added a comment to D18403: Move nvmecontrol to using linker sets.

I'll do the ones I marked and call it good and do round 2 for the const changes and the dlopen feature.

Sun, Dec 2, 7:01 PM
imp added a comment to D18403: Move nvmecontrol to using linker sets.

I've updated to be close to the end game. All that remains is readdir(/libexec/nvmecontrol and /usr/local/libexec/nvmecontrol) for all the .so's at startup and we'll be to the point where Netflix can deploy the Sekret Soss modules we have from Vendor X w/o changing base or polluting base with NDA material.

Sun, Dec 2, 6:58 PM
imp updated the diff for D18403: Move nvmecontrol to using linker sets.

Lots of cleanup, will be separate commits

Sun, Dec 2, 7:35 AM
imp added a comment to D18403: Move nvmecontrol to using linker sets.

Thanks for the feedback. I'll see how much I can include in my next round as some of them are asking for more structural changes beyond the scope of this set of changes (the suggestions are good, btw, just am keen to avoid too much scope creep).

Sun, Dec 2, 12:35 AM

Sat, Dec 1

imp added a comment to D18382: Add new "trim" conversion for dd(1).

There's a suggestion that the new trim program be named 'discard' to match linux's discard option on mount. Perhaps discard would be a better name. Though I hesitate to make such a bike-sheddy suggestion.

Sat, Dec 1, 9:58 PM
imp added reviewers for D18403: Move nvmecontrol to using linker sets: jimharris, cem.
Sat, Dec 1, 9:46 PM
imp updated the diff for D18403: Move nvmecontrol to using linker sets.

tweaks

Sat, Dec 1, 9:45 PM
imp created D18403: Move nvmecontrol to using linker sets.
Sat, Dec 1, 8:47 PM
imp added a comment to D18380: Fix a number of style issues with trim(8).

I'm pissed this was committed. It wasn't ready and in total breach of protocol.
This matter is *NOT* settled and you're lucky I don't just remove it from the tree.

Sat, Dec 1, 3:05 AM

Fri, Nov 30

imp added a comment to D18351: Add BIO_SPEEDUP.
In D18351#391315, @kib wrote:

And I still want my knob to not loose any trim :(.

Fri, Nov 30, 7:55 PM
imp accepted D16730: Committers Guide: Add a section encouraging pre-commit review..
Fri, Nov 30, 4:30 PM
imp added a comment to D17131: loader: create bio_alloc and bio_free for bios bounce buffer.

Couple of minor nits, but otherwise looks OK.

Fri, Nov 30, 4:16 PM
imp accepted D18382: Add new "trim" conversion for dd(1).

My acceptance is provisional here.

Fri, Nov 30, 2:29 PM
imp added inline comments to D18351: Add BIO_SPEEDUP.
Fri, Nov 30, 7:07 AM
imp added a comment to D18351: Add BIO_SPEEDUP.

This now looks reasonable to me.

At the moment during block shortages we request speedup of retiring TRIMs. But if there are no TRIMs (or we are on a disk that does not support TRIMs) then it seems to me that we should fall back to doing write speedup. Is this a change that we should make at the filesystem level, or should the drive recognize that if it is asked to speedup TRIMs and there are no TRIMs to speed up, then it should fall back to speeding up writes?

Fri, Nov 30, 7:07 AM

Thu, Nov 29

imp added a comment to D18382: Add new "trim" conversion for dd(1).

I like this, independent of a standalone utility.

Thu, Nov 29, 7:28 PM
imp added a reviewer for D18380: Fix a number of style issues with trim(8): cem.
Thu, Nov 29, 5:54 PM
imp added a reviewer for D18380: Fix a number of style issues with trim(8): eugen_grosbein.net.
Thu, Nov 29, 5:50 PM
imp created D18380: Fix a number of style issues with trim(8).
Thu, Nov 29, 5:48 PM

Wed, Nov 28

imp added a comment to D17947: Rework UEFI ESP generation.

a quick scan doesn't reveal anything major.

Wed, Nov 28, 6:48 PM
imp updated the diff for D18351: Add BIO_SPEEDUP.

First pass at turning off write limiting when in speedup mode.

Wed, Nov 28, 6:02 PM
imp added a comment to D18351: Add BIO_SPEEDUP.

OK. I've updated this in a few ways:

Wed, Nov 28, 5:37 PM
imp updated the diff for D18351: Add BIO_SPEEDUP.

Fixup after rebase oops

Wed, Nov 28, 5:27 PM
imp updated the diff for D18351: Add BIO_SPEEDUP.

Update based on Kirk's comments and some thinking about the problem.

Wed, Nov 28, 5:20 PM
imp updated the diff for D18351: Add BIO_SPEEDUP.

Ignore write speedup requests.

Wed, Nov 28, 6:11 AM
imp updated the diff for D18351: Add BIO_SPEEDUP.

Fold Kirk's suggestions in, but only his diff, not the one to remove one of my calls.

Wed, Nov 28, 5:57 AM
imp added a comment to D18351: Add BIO_SPEEDUP.

Kirk sent me this (patch has been removed, but integrated into next update, but I've not yet internalized it enough to comment tonight):

Wed, Nov 28, 5:46 AM

Tue, Nov 27

imp added a comment to D18351: Add BIO_SPEEDUP.
In D18351#389989, @kib wrote:

I never saw a situation where SU workitems were the cause of the memory shortage. I understand that slow trims can cause filesystem write hiccups and that the change would somewhat lessen this.

Since, as noted above, the SU workitems shortage does not cause system instability, only performance problems, I would prefer that the abort of requested trims was controlled by an option, perhaps even per-mount. For instance, on my w/s I never have low memory condition but often do large metadata-intensive ops and I do not want to loose the useful hints which manage the SSDs lifespan.

Tue, Nov 27, 3:32 PM
imp updated the diff for D18351: Add BIO_SPEEDUP.

Again with the changes, this time including all the changes.

Tue, Nov 27, 3:15 PM
imp updated the diff for D18351: Add BIO_SPEEDUP.

Address review feedback.

Tue, Nov 27, 3:14 PM
imp created D18351: Add BIO_SPEEDUP.
Tue, Nov 27, 12:48 AM
imp committed rS341005: NVME trim clocking.
NVME trim clocking
Tue, Nov 27, 12:37 AM

Mon, Nov 26

imp committed rS341003: Minor tweaks to the formatting.
Minor tweaks to the formatting
Mon, Nov 26, 10:52 PM

Thu, Nov 22

imp added inline comments to D18036: Fix style nits in strings.
Thu, Nov 22, 1:12 AM

Tue, Nov 20

imp committed rS340664: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.
Ensure that all values of ns, us and ms work for {n,u,m}stosbt
Tue, Nov 20, 7:11 AM
imp closed D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.
Tue, Nov 20, 7:11 AM
imp added a comment to D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.

updated with kasserts enabled as much as I can...

Tue, Nov 20, 2:42 AM
imp updated the diff for D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.

KASSERT where possible...

Tue, Nov 20, 2:40 AM

Mon, Nov 19

imp updated the diff for D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.

cem's changes

Mon, Nov 19, 11:01 PM
imp added a comment to D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.

OK. I think that some of the other ways to spell stuff may be more obscure.

Mon, Nov 19, 11:00 PM
imp updated the summary of D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.
Mon, Nov 19, 9:42 PM
imp updated the test plan for D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.
Mon, Nov 19, 8:58 PM
imp created D18051: Ensure that all values of ns, us and ms work for {n,u,m}stosbt.
Mon, Nov 19, 8:55 PM

Fri, Nov 16

imp closed D3461: Base of autoloading stuff.

This was committed a long time ago...

Fri, Nov 16, 10:03 PM
imp added a comment to D17705: Suggest upload of dmesg in login fortune file.

I don't think we need email= anymore.

Fri, Nov 16, 5:43 PM
imp committed rS340481: Remove do-nothing nvme_modevent..
Remove do-nothing nvme_modevent.
Fri, Nov 16, 4:51 PM

Thu, Nov 15

imp committed rS340452: Introduce scsi_ata_setfeatures() as a convenient way to make.
Introduce scsi_ata_setfeatures() as a convenient way to make
Thu, Nov 15, 4:02 PM
imp committed rS340453: Add cam_iosched_set_latfcn to set a latency callback for high latency..
Add cam_iosched_set_latfcn to set a latency callback for high latency.
Thu, Nov 15, 4:02 PM
imp committed rS340450: When converting ns,us,ms to sbt, return the ceil() of the result.
When converting ns,us,ms to sbt, return the ceil() of the result
Thu, Nov 15, 4:02 PM
imp committed rS340451: Do proper conversion to/from sbt..
Do proper conversion to/from sbt.
Thu, Nov 15, 4:02 PM

Wed, Nov 14

imp committed rS340448: Remove trailing white space in advance of other changes..
Remove trailing white space in advance of other changes.
Wed, Nov 14, 11:16 PM