jtl (Jonathan T. Looney)Administrator
User

Projects

User Details

User Since
Oct 29 2015, 5:25 PM (138 w, 3 h)
Roles
Administrator

Recent Activity

Mon, Jun 18

jtl added a comment to D15706: socket: introduction of destructor callback.
In D15706#334711, @jtl wrote:

Didn't quite catch this before it was committed. This isn't really a destructor, it's a close notification. Rather than confuse matters, as sockets have UMA destructors as well, this should probably be so_notify_close.

Yes, I caught that when I asked the submitter about the placement of the "destructor". But, once I figured out it was a close notification, I should have changed the name. Mea cupla.

Before changing this, let me see if I can confirm what the Linux implementation does.

Thanks, it is intended to be equivalent to sk_destruct() callback in struct sock in Linux.

Is the Linux sk_destruct called on socket close, or on socket destruction? If we want an actual socket destructor, we can add one of those [instead / as well], called from the socket destructor function.

It is literally destructor, which is called when the final reference count drops to zero. So probably we should define so_dtor() as destructor (so keep the name) and call it after SOCK_UNLOCK() in sofree(). (in Linux it is invoked without locking socket).

Mon, Jun 18, 11:07 PM
jtl accepted D15686: convert inpcbhash rlock to epoch.
Mon, Jun 18, 10:13 PM

Fri, Jun 15

jtl added a comment to D15706: socket: introduction of destructor callback.

Didn't quite catch this before it was committed. This isn't really a destructor, it's a close notification. Rather than confuse matters, as sockets have UMA destructors as well, this should probably be so_notify_close.

Fri, Jun 15, 6:19 PM
jtl updated subscribers of D15686: convert inpcbhash rlock to epoch.

I've spent some time thinking about this a bit, and I have the following comments. (Some may seem contradictory, but please bear with me. :-) )

Fri, Jun 15, 3:32 PM
jtl added inline comments to D15815: Set page permissions for kernel modules.
Fri, Jun 15, 1:59 AM

Thu, Jun 14

jtl created D15815: Set page permissions for kernel modules.
Thu, Jun 14, 11:27 PM

Wed, Jun 13

jtl committed rS335068: Make UMA and malloc(9) return non-executable memory in most cases..
Make UMA and malloc(9) return non-executable memory in most cases.
Wed, Jun 13, 5:05 PM
jtl closed D15691: Make kernel allocations be non-executable on some platforms.
Wed, Jun 13, 5:04 PM
jtl updated the diff for D15691: Make kernel allocations be non-executable on some platforms.

Address @markj's feedback by always defining the vmd_kernel_rwx_arena member of the vm_domain struct.

Wed, Jun 13, 4:47 PM
jtl added inline comments to D15691: Make kernel allocations be non-executable on some platforms.
Wed, Jun 13, 4:41 PM

Tue, Jun 12

jtl updated the diff for D15691: Make kernel allocations be non-executable on some platforms.

Update the zone(9) manpage.

Tue, Jun 12, 12:42 PM

Mon, Jun 11

jtl updated the diff for D15691: Make kernel allocations be non-executable on some platforms.

Okay, really without the cruft this time. (Hopefully...)

Mon, Jun 11, 11:55 PM
jtl added a comment to D15686: convert inpcbhash rlock to epoch.

I'll try to look at this review this week.

Mon, Jun 11, 11:54 PM
jtl added a reviewer for D15686: convert inpcbhash rlock to epoch: rwatson.
Mon, Jun 11, 11:53 PM
jtl updated the diff for D15691: Make kernel allocations be non-executable on some platforms.

Get the correct version of sys/vm/vm_kern.c (with the troubleshooting stuff removed).

Mon, Jun 11, 11:47 PM
jtl updated the diff for D15691: Make kernel allocations be non-executable on some platforms.

Numerous updates:

  • Address @alc's concerns by adding a new arena for allocations with non-standard permissions. Import a 2MB-aligned address block at a time into the arena. Release space back to the parent arena when able. But, only do this for architectures with superpages.
  • Plumb M_EXEC through malloc(9).
  • Fix BPF by reverting most of rS317072.
  • Add a note to the manpage that not all architectures will enforce execution permissions.
Mon, Jun 11, 11:38 PM
jtl committed rS334983: Fix a memory leak for the BIOCSETWF ioctl on kernels with the BPF_JITTER.
Fix a memory leak for the BIOCSETWF ioctl on kernels with the BPF_JITTER
Mon, Jun 11, 11:32 PM
jtl added a comment to D15758: Rack crash that Larry Rosenman found...

This passes my "sniff test", but it would be better to get @pkelsey to review it.

Mon, Jun 11, 2:49 PM
jtl added reviewers for D15758: Rack crash that Larry Rosenman found..: transport, pkelsey.
Mon, Jun 11, 2:45 PM
jtl requested changes to D15757: HPTS Is missing some support for Vnets.

I think there are more changes needed.

Mon, Jun 11, 2:37 PM
jtl closed D15756: Change RACK dependency on TCPHPTS from build-time to load-time.
Mon, Jun 11, 2:27 PM
jtl committed rS334949: Change RACK dependency on TCPHPTS from a build-time dependency to a load-.
Change RACK dependency on TCPHPTS from a build-time dependency to a load-
Mon, Jun 11, 2:27 PM
jtl accepted D15634: Use updated TCP timestamps when retransmitting SYN-ACK using the syncache code path.
Mon, Jun 11, 10:36 AM
jtl created D15756: Change RACK dependency on TCPHPTS from build-time to load-time.
Mon, Jun 11, 10:31 AM

Fri, Jun 8

jtl committed rS334854: Create a symlink for sodtor_set(9) to the socket(9) man page..
Create a symlink for sodtor_set(9) to the socket(9) man page.
Fri, Jun 8, 7:47 PM
jtl closed D15706: socket: introduction of destructor callback.
Fri, Jun 8, 7:35 PM
jtl committed rS334853: Add a socket destructor callback. This allows kernel providers to set.
Add a socket destructor callback. This allows kernel providers to set
Fri, Jun 8, 7:35 PM
jtl accepted D15706: socket: introduction of destructor callback.

The submitter spoke to me in person at BSDCan and answered my questions.

Fri, Jun 8, 6:38 PM
jtl added a comment to D15691: Make kernel allocations be non-executable on some platforms.
In D15691#331801, @alc wrote:

Overall, I think that this is a good idea, but the implementation has the following problem. The allocation of one executable page will block the promotion of the surrounding pages to a superpage mapping.

Fri, Jun 8, 3:57 PM
jtl added inline comments to D15706: socket: introduction of destructor callback.
Fri, Jun 8, 3:26 PM
jtl added a reviewer for D15706: socket: introduction of destructor callback: network.
Fri, Jun 8, 3:26 PM

Thu, Jun 7

jtl updated the diff for D15691: Make kernel allocations be non-executable on some platforms.

Incorporate review feedback from @jhb.

Thu, Jun 7, 4:53 PM
jtl created D15691: Make kernel allocations be non-executable on some platforms.
Thu, Jun 7, 4:04 PM
jtl committed rS334783: Fix a typo in vm_domain_set(). When a domain crosses into the severe range,.
Fix a typo in vm_domain_set(). When a domain crosses into the severe range,
Thu, Jun 7, 1:32 PM
jtl added a comment to D15686: convert inpcbhash rlock to epoch.

I added a few basic comments while I ponder the rest...

Thu, Jun 7, 2:46 AM
jtl added a reviewer for D15686: convert inpcbhash rlock to epoch: glebius.
Thu, Jun 7, 2:28 AM

Fri, Jun 1

jtl added a comment to D15646: Provide option to panic when the IPMI creates an NMI.

I don't know ISA well. I'm open to switching this to be an option to always panic on any NMI, instead of picking NMI_TIMER2 for special treatment.

Fri, Jun 1, 5:33 PM
jtl created D15646: Provide option to panic when the IPMI creates an NMI.
Fri, Jun 1, 5:05 PM
jtl created D15645: Update atomic man page to reflect reality.
Fri, Jun 1, 4:52 PM
jtl committed rS334490: Update the sysctl(9) manpage to indicate that <sys/param.h> is required.
Update the sysctl(9) manpage to indicate that <sys/param.h> is required
Fri, Jun 1, 4:47 PM

Thu, May 31

jtl accepted D15222: Use UDP len when calculating UDP checksums.

I'm sorry it took me so long to review this.

Thu, May 31, 3:08 PM
jtl added a comment to D15386: Improve portability of structs xinpcb and friends..

Anyone on transport care about this?

Thu, May 31, 1:26 AM

Thu, May 24

jtl added a comment to D11003: Load balance sockets with new SO_REUSEPORT_LB option.
In D11003#327485, @jtl wrote:

Is everyone on this review fine with me committing this? With the inheritance code removed, I believe there are no further objections or questions.

I've been meaning to review this, but have been delayed. I would appreciate a few extra days to look it over.

@jtl I'd like to commit this at some point this week, but I also want your feedback on it. Should we hold off until next week to drop this into head?

Thu, May 24, 12:43 AM

May 21 2018

jtl added a comment to D11003: Load balance sockets with new SO_REUSEPORT_LB option.

Is everyone on this review fine with me committing this? With the inheritance code removed, I believe there are no further objections or questions.

May 21 2018, 9:09 PM
jtl accepted D15510: Defer inpcb deletion until after a grace period has elapsed.

Having stared at this a bit, I generally think its fine. In fact, in some aspects it is an improvement over what it replaces.

May 21 2018, 6:59 PM
jtl added a reviewer for D15510: Defer inpcb deletion until after a grace period has elapsed: rwatson.
May 21 2018, 6:21 PM
jtl added a comment to D15483: More bcmp "optimization".
In D15483#326763, @mjg wrote:

First a minor note is that I took the liberty of s/bzero/bcmp, which I presume was intended.

May 21 2018, 2:48 PM

May 18 2018

jtl added a comment to D15483: More bcmp "optimization".

By the way, measurements were taken on an Intel E5-2697A v4 (32-core Broadwell).

May 18 2018, 7:26 PM
jtl created D15483: More bcmp "optimization".
May 18 2018, 7:18 PM

May 16 2018

jtl added a comment to D7135: A problem with ASYNC drain.

Yes, see rS302894, rS302997, rS302998, rS303037, rS303096, and rS303098.

May 16 2018, 6:49 PM

May 13 2018

jtl added a comment to D15222: Use UDP len when calculating UDP checksums.

Sorry it took me a while to look at this. See comments in-line.

May 13 2018, 12:32 AM
jtl added a reviewer for D15222: Use UDP len when calculating UDP checksums: transport.
May 13 2018, 12:31 AM

May 12 2018

jtl closed D15021: Wakup applications with shutdown on listen sockets (stable/11 version).
May 12 2018, 1:55 AM
jtl committed rS333511: r285910 attempted to make shutdown() be POSIX compliant by returning.
r285910 attempted to make shutdown() be POSIX compliant by returning
May 12 2018, 1:55 AM

May 11 2018

jtl added a comment to D15021: Wakup applications with shutdown on listen sockets (stable/11 version).

Given the fact that users appear to depend on this (we have two PRs on it), this appears to have been an unintended change in stable/11, and it seems to violate POLA by making this change on a stable branch, I intend to commit this today unless someone objects.

May 11 2018, 1:19 PM

May 10 2018

jtl added a comment to D15019: Wakup applications with shutdown on listen sockets.

@glebius : Any progress on reviewing this? IIRC, you didn't like overloading the SS_ISDISCONNECTED flag to indicate listen sockets that were shutdown. Do you have an alternate proposal?

May 10 2018, 7:06 PM
jtl added a comment to D15021: Wakup applications with shutdown on listen sockets (stable/11 version).

Is everyone OK with this going in? Or, to ask it differently, is anyone not OK? The 11.2 release is underway, so it would be good to get this in soon, if we agree to fix it in stable/11.

May 10 2018, 7:04 PM

May 9 2018

jtl added a comment to D15358: Address memory leak in new reno cc module.

BTW, it looks like this was introduced in r331214. Since it is so new, my "not widely used" comment is probably both true and somewhat irrelevant.

May 9 2018, 12:09 AM
jtl added inline comments to D15358: Address memory leak in new reno cc module.
May 9 2018, 12:03 AM

May 8 2018

jtl added reviewers for D15358: Address memory leak in new reno cc module: transport, rrs.
May 8 2018, 10:32 PM

May 7 2018

jtl added a comment to D15337: Add support for higher resolution timestamps.
In D15337#323183, @jtl wrote:

How does this interact with the low-latency, high-precision timestamp option being discussed at the IETF?

See https://tools.ietf.org/html/draft-wang-tcpm-low-latency-opt-00 and https://www.ietf.org/proceedings/97/slides/slides-97-tcpm-tcp-options-for-low-latency-00.pdf .

@jtl I was under the impression that that discussion had stalled. The RFC you point at expired on December 10, 2017. Is there a mailing list where this is being discussed or do I need to mail them individually?

May 7 2018, 6:26 PM
jtl added inline comments to D11003: Load balance sockets with new SO_REUSEPORT_LB option.
May 7 2018, 3:30 PM
jtl added a comment to D15337: Add support for higher resolution timestamps.

How does this interact with the low-latency, high-precision timestamp option being discussed at the IETF?

May 7 2018, 2:17 PM

Apr 27 2018

jtl added a comment to D14993: Add a limit on how long TCP data can live at the head of the output queue.

I'm not sure it is a good idea to have a sysctl for the default value. I would prefer to have the default value always be 0 (not changeable by a sysctl) and required the application to set a non-default value via the socket option. Why is there a need to change the default from "off" on a system wide base?

Apr 27 2018, 12:26 PM

Apr 25 2018

jtl created D15203: Post-panic KASSERT enhancements.
Apr 25 2018, 8:27 PM

Apr 24 2018

jtl committed rS332958: Update r332860 by changing the default from suppressing post-panic.
Update r332860 by changing the default from suppressing post-panic
Apr 24 2018, 6:47 PM

Apr 23 2018

jtl committed rS332889: MFC r331745 (by np):.
MFC r331745 (by np):
Apr 23 2018, 2:22 PM

Apr 21 2018

jtl committed rS332860: When running with INVARIANTS, the kernel contains extra checks. However,.
When running with INVARIANTS, the kernel contains extra checks. However,
Apr 21 2018, 5:05 PM
jtl closed D12920: Avoid a double-panic caused by KASSERTs.
Apr 21 2018, 5:05 PM

Apr 20 2018

jtl committed rS332842: MFC r319216:.
MFC r319216:
Apr 20 2018, 8:18 PM
jtl committed rS332841: MFC r319215:.
MFC r319215:
Apr 20 2018, 8:17 PM
jtl committed rS332840: MFC r319214:.
MFC r319214:
Apr 20 2018, 8:10 PM
jtl added a comment to D15021: Wakup applications with shutdown on listen sockets (stable/11 version).

Presumably we need to agree on and commit D15019 first, and commit this as a MFC in spirit after the change in HEAD.

Apr 20 2018, 6:44 PM
jtl committed rS332834: MFC r314286:.
MFC r314286:
Apr 20 2018, 6:37 PM
jtl committed rS332831: MFC r314116:.
MFC r314116:
Apr 20 2018, 3:55 PM
jtl committed rS332830: MFC r313447:.
MFC r313447:
Apr 20 2018, 3:49 PM
jtl committed rS332829: MFC r307083:.
MFC r307083:
Apr 20 2018, 3:44 PM
jtl committed rS332828: MFC r306768:.
MFC r306768:
Apr 20 2018, 3:41 PM
jtl committed rS332827: MFC r330511:.
MFC r330511:
Apr 20 2018, 3:36 PM
jtl committed rS332826: MFC r330510:.
MFC r330510:
Apr 20 2018, 3:29 PM
jtl committed rS332823: MFC r329171:.
MFC r329171:
Apr 20 2018, 3:06 PM
jtl committed rS332822: MFC r329071:.
MFC r329071:
Apr 20 2018, 2:55 PM
jtl committed rS332821: MFC r331309:.
MFC r331309:
Apr 20 2018, 2:47 PM
jtl committed rS332820: MFC r331484:.
MFC r331484:
Apr 20 2018, 2:34 PM
jtl committed rS332819: MFC r331488:.
MFC r331488:
Apr 20 2018, 2:21 PM
jtl committed rS332818: MFC r331926:.
MFC r331926:
Apr 20 2018, 2:01 PM
jtl committed rS332817: MFC r332120:.
MFC r332120:
Apr 20 2018, 1:59 PM
jtl added a comment to D15021: Wakup applications with shutdown on listen sockets (stable/11 version).

If we're going to get this into 11.2, it would be nice if someone reviewed this soonish.

Apr 20 2018, 1:46 PM

Apr 11 2018

jtl accepted D15043: Add Tom Jones (thj) as a new src committer..
Apr 11 2018, 10:54 PM
jtl accepted D15046: Make lagg creation more fault tolerant.
Apr 11 2018, 5:56 PM
jtl accepted D15043: Add Tom Jones (thj) as a new src committer..

Looks good. Please check on two things before committing:

  1. I don't see a svn:keywords property change for head/share/pgpkeys/thj.key. I expected to see one, although this could just be a Phabricator display issue. Can you confirm you've followed the instructions in the Note at the start of 5.3.7 of the Committer's Guide ("Adding and Removing Files")? If not, please do so and then svn remove and svn add the file again. You can check that the property was correctly applied by running this command: svn propget svn:keywords thj.key. You should see FreeBSD=%H.
  2. Please check on the indentation in news.xml. (See in-line note.)
Apr 11 2018, 5:45 PM
jtl requested changes to D15043: Add Tom Jones (thj) as a new src committer..
Apr 11 2018, 5:07 PM
jtl accepted D15042: Add myself to committers-src.dot.
Apr 11 2018, 4:55 PM

Apr 10 2018

jtl added inline comments to D15020: Add of high-precision-timer-system used by rack and bbr for pacing. Also sync up the function block differences needed by bbr/rack..
Apr 10 2018, 10:45 PM
jtl accepted D15020: Add of high-precision-timer-system used by rack and bbr for pacing. Also sync up the function block differences needed by bbr/rack..

See nits inline.

Apr 10 2018, 6:41 PM
jtl committed rS332382: Add missing header change from r332381..
Add missing header change from r332381.
Apr 10 2018, 5:00 PM
jtl committed rS332381: Modify the net.inet.tcp.function_ids sysctl introduced in r331347..
Modify the net.inet.tcp.function_ids sysctl introduced in r331347.
Apr 10 2018, 4:59 PM
jtl committed rS332379: Move the TCP Blackbox Recorder probe in tcp_output.c to be with the.
Move the TCP Blackbox Recorder probe in tcp_output.c to be with the
Apr 10 2018, 3:54 PM
jtl committed rS332378: Clean up some debugging code left in tcp_log_buf.c from r331347..
Clean up some debugging code left in tcp_log_buf.c from r331347.
Apr 10 2018, 3:52 PM

Apr 9 2018

jtl committed rS332340: Add Tom Jones (thj@) as a source committer..
Add Tom Jones (thj@) as a source committer.
Apr 9 2018, 9:47 PM