Re: Kernel compilation fail with clang.

On Wed, Aug 02, 2017 at 09:06:00PM +0200, Olivier Antoine wrote: > Hi, > > I may have missed something but I'm unable to compile the kernel since (I > think) the switch to clang on AMD64. > > I have upgraded to the latest snapshot, updated my /sys tree > > $ cd

Remove duplicated commented out include in files.i386

Hi tech@, Remove duplicated commented out "dev/rasops/files.rasops" include. Comments? OK? Index: sys/arch/i386/conf/files.i386 === RCS file: /cvs/src/sys/arch/i386/conf/files.i386,v retrieving revision 1.233 diff -u -p -r1.233

Re: less(1) - segmentation fault with '-g'

On Wed, Aug 02, 2017 at 08:46:26PM +0200, Martijn van Duren wrote: > You're right. Maybe I should read up on my less. :-) > > OK martijn@ for the full diff. Thanks, committed. > On 08/02/17 19:55, Anton Lindqvist wrote: > > On Wed, Aug 02, 2017 at 07:19:50PM +0200, Martijn van Duren wrote: > >>

Kernel compilation fail with clang.

Hi, I may have missed something but I'm unable to compile the kernel since (I think) the switch to clang on AMD64. I have upgraded to the latest snapshot, updated my /sys tree $ cd /sys/arch/$(machine)/compile/GENERIC.MP $ doas make obj $ make config $ make Then the compilation fail with this

Fix libexec/ld.so/dlclose regress

Couldn't convince clang not to inline duplicateFun() into bbTest2(). Splitting things out in a seperate file avoids the issue. Fixes the regression test. ok? Index: regress/libexec/ld.so/dlclose/test1/libbb/Makefile === RCS file:

Fix libunwind on i386

So apparently Apple messed up the register numbering, or at least chose to be different: https://reviews.llvm.org/D22508 Fixes the exception handling regression tests on i386 with clang. ok? Index: lib/libunwind/include/libunwind.h

Re: less(1) - segmentation fault with '-g'

You're right. Maybe I should read up on my less. :-) OK martijn@ for the full diff. On 08/02/17 19:55, Anton Lindqvist wrote: > On Wed, Aug 02, 2017 at 07:19:50PM +0200, Martijn van Duren wrote: >> OK martijn@ for the pattern.c part. >> >> The search.c part seems unneeded. >> - If we don't have

Re: make pkg_info -Q work with other flags

On Sat, Jul 29, 2017 at 10:40:37AM -0600, Aaron Bieber wrote: > Hola, > > Currently "pkg_info -Q" doesn't respect other flags and the way > pkg_info(1) reads, it implies that they will work with it. > > This diff makes pkg_info function as expected when other flags are > passed when using -Q.

Re: less(1) - segmentation fault with '-g'

On Wed, Aug 02, 2017 at 07:19:50PM +0200, Martijn van Duren wrote: > OK martijn@ for the pattern.c part. > > The search.c part seems unneeded. > - If we don't have a match we don't enter hilite_line at all. > - If we have a match and don't negate get sp/ep for every match. > - If we have a match

Re: less(1) - segmentation fault with '-g'

OK martijn@ for the pattern.c part. The search.c part seems unneeded. - If we don't have a match we don't enter hilite_line at all. - If we have a match and don't negate get sp/ep for every match. - If we have a match and negate, with your process.c patch we get a sp/ep with value NULL and return

Re: fstat output

On Wed, 02 Aug 2017 11:01:38 +0200, Alexander Bluhm wrote: > On Wed, Aug 02, 2017 at 09:56:50AM +0200, Martin Pieuchot wrote: > > Simple diff to improved readability. Before: > > There are situations where you want a full match on the command. > For testing and scripting with long program names

Re: clang ld.so regress failures

On Wed, Aug 02, 2017 at 12:09:13AM -0400, Ted Unangst wrote: > Mark Kettenis wrote: > > FAIL libexec/ld.so/dlclose/test1/prog3/prog3 > > > > This fails because clang doesn't respect ELF interposition: > > > > http://lists.llvm.org/pipermail/llvm-dev/2016-November/107625.html > > > > We

nd6 address expiration & NET_LOCK() contention

Florian killed the default router and prefix lists from the kernel. So nd6_timer() is now only doing address expiration based on pltime/vltime. Diff below renames the function and kill outdated comments to reflect the reality. Now, this timer wakes a task every second. And even if there's

Re: clang ld.so regress failures

> From: "Ted Unangst" > Date: Wed, 02 Aug 2017 00:09:13 -0400 > > Mark Kettenis wrote: > > FAIL libexec/ld.so/dlclose/test1/prog3/prog3 > > > > This fails because clang doesn't respect ELF interposition: > > > >

Re: Reducing NET_LOCK() contention

On 2.8.2017. 11:00, Alexandr Nedvedicky wrote: > Hello, > > On Wed, Aug 02, 2017 at 10:10:51AM +0200, Martin Pieuchot wrote: >> On 18/07/17(Tue) 15:55, Martin Pieuchot wrote: >>> When forwarding a lot of traffic with 10G interfaces contention on the >>> NET_LOCK() is "visible". Each time you

Re: Reducing NET_LOCK() contention

On 2.8.2017. 10:10, Martin Pieuchot wrote: > On 18/07/17(Tue) 15:55, Martin Pieuchot wrote: >> When forwarding a lot of traffic with 10G interfaces contention on the >> NET_LOCK() is "visible". Each time you type "ifconfig" you can go grab >> a coffee... >> >> The problem has a name:

so{s,g}etopt() & solock

Diff below moves the socket lock "above" sosetopt(), sogetopt() and sosplice(). While this adds a lot of lock/unlock dances in NFS, they will be merge in a later diff. sosetopt() modifies a socket fields so it needs the lock. sogetopt() do not always need it, but it makes the code simpler to

Re: Reducing NET_LOCK() contention

Hello, On Wed, Aug 02, 2017 at 10:10:51AM +0200, Martin Pieuchot wrote: > On 18/07/17(Tue) 15:55, Martin Pieuchot wrote: > > When forwarding a lot of traffic with 10G interfaces contention on the > > NET_LOCK() is "visible". Each time you type "ifconfig" you can go grab > > a coffee... > > > >

Re: fstat output

On Wed, Aug 02, 2017 at 09:56:50AM +0200, Martin Pieuchot wrote: > Simple diff to improved readability. Before: There are situations where you want a full match on the command. For testing and scripting with long program names this is necessary. With a -v switch that shows the long name I would

Re: em: workaround for ultra-low-power mode on i219V

On Tue, 1 Aug 2017, Stefan Fritsch wrote: > For consumer NICs, the patch only does something on I218_LM_3/I218_V_3 or > I219* or newer. Sorry, misread the code there. There are more i218 variants that are affected by the patch.

Re: Reducing NET_LOCK() contention

On 18/07/17(Tue) 15:55, Martin Pieuchot wrote: > When forwarding a lot of traffic with 10G interfaces contention on the > NET_LOCK() is "visible". Each time you type "ifconfig" you can go grab > a coffee... > > The problem has a name: pf_purge_thread(). This thread is created by > default and

fstat output

Simple diff to improved readability. Before: $ fstat -f /tmp USER CMD PID FD MOUNTINUM MODE R/WSZ|DV mpi fstat 964791 /tmp 19 -rw-r--r-- w0 mpi ksh466571 /tmp 19 -rw-r--r-- w0 mpi