Page MenuHomeFreeBSD
Feed Advanced Search

Today

hselasky committed rS352481: MFC r352137:.
MFC r352137:
Wed, Sep 18, 7:32 AM
hselasky committed rS352480: MFC r352137:.
MFC r352137:
Wed, Sep 18, 7:31 AM
hselasky committed rS352479: MFC r352207:.
MFC r352207:
Wed, Sep 18, 7:29 AM
hselasky committed rS352478: MFC r352207:.
MFC r352207:
Wed, Sep 18, 7:29 AM
hselasky committed rS352477: MFC r352206:.
MFC r352206:
Wed, Sep 18, 7:26 AM
hselasky committed rS352476: MFC r352206:.
MFC r352206:
Wed, Sep 18, 7:26 AM
hselasky committed rS352474: MFC r352205:.
MFC r352205:
Wed, Sep 18, 7:24 AM
hselasky committed rS352472: MFC r352205:.
MFC r352205:
Wed, Sep 18, 7:21 AM

Yesterday

hselasky accepted D21527: Add SIOCGIFDOWNREASON.
Tue, Sep 17, 1:12 PM

Sat, Sep 14

hselasky committed rS352330: MFC r351693:.
MFC r351693:
Sat, Sep 14, 1:34 PM
hselasky committed rS352328: MFC r351693:.
MFC r351693:
Sat, Sep 14, 1:33 PM
hselasky committed rS352326: MFC r351847:.
MFC r351847:
Sat, Sep 14, 1:31 PM
hselasky committed rS352324: MFC r351847:.
MFC r351847:
Sat, Sep 14, 1:30 PM
hselasky committed rP512003: Update qaudiosonar to v1.6.1..
Update qaudiosonar to v1.6.1.
Sat, Sep 14, 10:04 AM

Fri, Sep 13

hselasky abandoned D21617: Improve support for refcount waiting.

I will use the completion() API in the LinuxKPI for my purpose instead. Discussed with @mjg .

Fri, Sep 13, 4:43 PM
hselasky abandoned D21621: Implement m_snd_tag_wait().

refcount_wait() is expected to go away.

Fri, Sep 13, 4:40 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Fri, Sep 13, 4:25 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Fri, Sep 13, 4:13 PM
hselasky added a comment to D21617: Improve support for refcount waiting.

@mjg: what about the addition of refcount_releasen_if_not_last()? Is that part OK?

Fri, Sep 13, 4:12 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Style and comment fixes from @kib

Fri, Sep 13, 7:19 AM
hselasky added inline comments to D21621: Implement m_snd_tag_wait().
Fri, Sep 13, 7:12 AM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Implement @kib 's suggestion.

Fri, Sep 13, 7:00 AM

Thu, Sep 12

hselasky updated the diff for D21621: Implement m_snd_tag_wait().

Update as per Jeff's suggestion.

Thu, Sep 12, 9:57 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Thu, Sep 12, 9:49 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Update code as per Jeff's suggestions.

Thu, Sep 12, 9:49 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Update patch as per @jeff 's suggestion.

Thu, Sep 12, 7:40 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Thu, Sep 12, 7:34 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Thu, Sep 12, 7:27 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Thu, Sep 12, 7:19 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Catch case where refcount is re-incremented from zero. Noted by mjg@.

Thu, Sep 12, 7:05 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Thu, Sep 12, 6:50 PM
hselasky added inline comments to D21617: Improve support for refcount waiting.
Thu, Sep 12, 6:48 PM
hselasky created D21621: Implement m_snd_tag_wait().
Thu, Sep 12, 5:24 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Minor code refactor - lock sleep queue before dropping Giant. Makes code look better.

Thu, Sep 12, 4:55 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Rebase patch and use atomic_swap_int().

Thu, Sep 12, 4:51 PM
hselasky added a comment to D21617: Improve support for refcount waiting.

I can use atomic_swap() . Is that available for all platforms?

Thu, Sep 12, 4:34 PM
hselasky committed rS352253: Use REFCOUNT_COUNT() to obtain refcount where appropriate..
Use REFCOUNT_COUNT() to obtain refcount where appropriate.
Thu, Sep 12, 4:27 PM
hselasky closed D21620: Use REFCOUNT_COUNT() to obtain refcount where appropriate.
Thu, Sep 12, 4:27 PM
hselasky added a comment to D21620: Use REFCOUNT_COUNT() to obtain refcount where appropriate.

There is no observed bug.

Thu, Sep 12, 4:21 PM
hselasky committed rS352251: Use %u instead of %d to print unsigned integer..
Use %u instead of %d to print unsigned integer.
Thu, Sep 12, 4:20 PM
hselasky updated the diff for D21620: Use REFCOUNT_COUNT() to obtain refcount where appropriate.

Fix style.

Thu, Sep 12, 4:06 PM
hselasky committed rS352250: Fix for undefined behaviour..
Fix for undefined behaviour.
Thu, Sep 12, 3:58 PM
hselasky created D21620: Use REFCOUNT_COUNT() to obtain refcount where appropriate.
Thu, Sep 12, 3:46 PM
hselasky added a reviewer for D21617: Improve support for refcount waiting: markj.
Thu, Sep 12, 3:36 PM
hselasky updated the summary of D21617: Improve support for refcount waiting.
Thu, Sep 12, 2:59 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Keep API compatibility.

Thu, Sep 12, 2:58 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Add missed _refcount_update_saturated() call.

Thu, Sep 12, 2:33 PM
hselasky updated the diff for D21617: Improve support for refcount waiting.

Use "const u_int n" for input "n".

Thu, Sep 12, 2:27 PM
hselasky created D21617: Improve support for refcount waiting.
Thu, Sep 12, 2:26 PM
hselasky added inline comments to D21610: EPOCH tracing support for debugging..
Thu, Sep 12, 6:34 AM

Wed, Sep 11

hselasky added inline comments to D21527: Add SIOCGIFDOWNREASON.
Wed, Sep 11, 8:34 AM
hselasky committed rS352207: Use true and false when dealing with bool type in the LinuxKPI..
Use true and false when dealing with bool type in the LinuxKPI.
Wed, Sep 11, 8:25 AM
hselasky committed rS352206: Fix synchronous work drain issue in the LinuxKPI..
Fix synchronous work drain issue in the LinuxKPI.
Wed, Sep 11, 8:20 AM
hselasky committed rS352205: Fix broken DECLARE_TASKLET() macro after r347852..
Fix broken DECLARE_TASKLET() macro after r347852.
Wed, Sep 11, 7:53 AM
hselasky added inline comments to D21527: Add SIOCGIFDOWNREASON.
Wed, Sep 11, 7:42 AM

Tue, Sep 10

hselasky committed rS352142: Compile fixes for LINT..
Compile fixes for LINT.
Tue, Sep 10, 4:35 PM
hselasky committed rS352139: Merge ^/head r346083 through r352137..
Merge ^/head r346083 through r352137.
Tue, Sep 10, 3:51 PM
hselasky committed rS352137: Callout drain does not have to be followed by a callout stop call..
Callout drain does not have to be followed by a callout stop call.
Tue, Sep 10, 2:33 PM

Mon, Sep 9

hselasky added a comment to D21527: Add SIOCGIFDOWNREASON.

What about translation of text messages? Should we use an enum for this? Or provide both?

Mon, Sep 9, 8:56 AM

Fri, Sep 6

hselasky accepted D21542: LinuxKPI: Improve sysfs support..

Make sure:
make buildkernel WITH_OFED=YES

Fri, Sep 6, 3:07 PM
hselasky added inline comments to D21542: LinuxKPI: Improve sysfs support..
Fri, Sep 6, 6:27 AM

Thu, Sep 5

hselasky committed rP511175: Update webcamd to version 5.3.7.0.
Update webcamd to version 5.3.7.0
Thu, Sep 5, 11:20 AM
hselasky committed rS351847: Decrease the default audio playback latency to a maximum of 21.3ms..
Decrease the default audio playback latency to a maximum of 21.3ms.
Thu, Sep 5, 10:49 AM
hselasky committed rS351845: MFC r351167:.
MFC r351167:
Thu, Sep 5, 9:35 AM
hselasky committed rS351844: MFC r351167:.
MFC r351167:
Thu, Sep 5, 9:35 AM
hselasky committed rS351843: MFC r351146:.
MFC r351146:
Thu, Sep 5, 9:34 AM
hselasky committed rS351842: MFC r351146:.
MFC r351146:
Thu, Sep 5, 9:32 AM
hselasky committed rS351841: MFC r351009:.
MFC r351009:
Thu, Sep 5, 9:31 AM
hselasky committed rS351840: MFC r351009:.
MFC r351009:
Thu, Sep 5, 9:29 AM
hselasky committed rS351839: MFC r347387:.
MFC r347387:
Thu, Sep 5, 9:23 AM
hselasky committed rS351838: MFC r347387:.
MFC r347387:
Thu, Sep 5, 9:20 AM

Mon, Sep 2

hselasky added a comment to D17599: Fix for double free when deleting entries from epoch managed lists.

This patch likely needs to be rebased before we can continue.

Mon, Sep 2, 3:12 PM
hselasky added a comment to D17599: Fix for double free when deleting entries from epoch managed lists.

I don't see any panics due to this, so it is not a big issue.

Mon, Sep 2, 3:11 PM
hselasky committed rS351693: Use DEVICE memory instead of UNCACHEABLE on aarch64 in ioremap() in the….
Use DEVICE memory instead of UNCACHEABLE on aarch64 in ioremap() in the…
Mon, Sep 2, 8:34 AM
hselasky closed D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap.
Mon, Sep 2, 8:34 AM · arm64
hselasky committed rS351692: Fix regression issue after r351616. Make sure the mbuf queue gets initialized..
Fix regression issue after r351616. Make sure the mbuf queue gets initialized.
Mon, Sep 2, 8:31 AM

Sun, Sep 1

D21475: LinuxKPI: Add sysfs create/remove functions that handles multiple files in one call. is now accepted and ready to land.

Don't forget to MFC.

Sun, Sep 1, 5:20 PM

Fri, Aug 30

hselasky added a comment to D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap.

I'll try to get the patch in on Monday. I'm currently busy. Thank you for the patience.

Fri, Aug 30, 11:33 AM · arm64

Thu, Aug 29

hselasky accepted D21452: ichsmb: defer smbus attach until interrupts are available.
Thu, Aug 29, 7:16 AM
D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap is now accepted and ready to land.
Thu, Aug 29, 7:13 AM · arm64
D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap now requires review to proceed.
Thu, Aug 29, 7:13 AM · arm64
hselasky added a comment to D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap.

Just select commander revision in the action menu.

Thu, Aug 29, 7:12 AM · arm64

Wed, Aug 28

hselasky added a comment to D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap.

Can you update my patch and test an again? Thank you!

Wed, Aug 28, 7:28 PM · arm64
hselasky added a comment to D21452: ichsmb: defer smbus attach until interrupts are available.

If this driver doesn't need to started early have a look at this:

Wed, Aug 28, 1:59 PM
hselasky added a comment to D21452: ichsmb: defer smbus attach until interrupts are available.

Can you upload full context?

Wed, Aug 28, 1:50 PM
hselasky added inline comments to D21452: ichsmb: defer smbus attach until interrupts are available.
Wed, Aug 28, 1:48 PM
hselasky added inline comments to D21452: ichsmb: defer smbus attach until interrupts are available.
Wed, Aug 28, 1:47 PM
D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap is now accepted and ready to land.
Wed, Aug 28, 8:16 AM · arm64
hselasky added a comment to D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap.

Can you test this patch?

Wed, Aug 28, 8:15 AM · arm64
hselasky edited reviewers for D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap, added: greg_unrelenting.technology; removed: hselasky.
Wed, Aug 28, 8:14 AM · arm64
hselasky added a comment to D20789: Use DEVICE memory instead of UNCACHEABLE on aarch64 in LinuxKPI ioremap.

Pong :-) I'll have a look at the patch again.

Wed, Aug 28, 7:27 AM · arm64

Tue, Aug 27

hselasky added a comment to D21420: Fix logic in domainset_empty_vm()..

This fixes boot issues with DELL PowerEdge R620 BIOS v2.7.0.

Tue, Aug 27, 1:06 PM

Sun, Aug 25

hselasky accepted D21409: mixer(8): Report an error if passed value is an empty string.
Sun, Aug 25, 2:43 PM

Thu, Aug 22

hselasky added a comment to D21362: assert_mtx: treat LA_LOCKED as the same of LA_XLOCKED..

Why not extend the switch case in __mtx_assert ??

void
__mtx_assert(const volatile uintptr_t *c, int what, const char *file, int line)
{
        const struct mtx *m;
Thu, Aug 22, 7:30 AM

Wed, Aug 21

hselasky accepted D21277: Add kernel-side support for in-kernel TLS..
Wed, Aug 21, 9:45 AM

Aug 17 2019

hselasky added a comment to D21298: OFED: Unbreak SDP support in ibcore.

If all compilers support \#pragma once I'm fine with that, but else we should follow the style in FreeBSD.

Aug 17 2019, 10:03 AM
hselasky committed rS351167: Include item position in report descriptor dump in usbhidctl(1)..
Include item position in report descriptor dump in usbhidctl(1).
Aug 17 2019, 8:29 AM
D21298: OFED: Unbreak SDP support in ibcore is now accepted and ready to land.

Looks good. See my one comment.

Aug 17 2019, 8:15 AM

Aug 16 2019

hselasky committed rS351146: Implement detach_kernel_driver command in usbconfig(8)..
Implement detach_kernel_driver command in usbconfig(8).
Aug 16 2019, 9:18 PM
hselasky added inline comments to D21277: Add kernel-side support for in-kernel TLS..
Aug 16 2019, 11:15 AM