On Sat, Oct 5, 2013 at 5:52 AM, Mark Kettenis mark.kette...@xs4all.nl wrote:
Date: Thu, 3 Oct 2013 18:12:59 -0700
From: Philip Guenther guent...@gmail.com
...
@@ -196,11 +167,15 @@ hardclock(struct clockframe *frame)
*/
if (CLKF_USERMODE(frame
On Fri, 16 Aug 2013, Ted Unangst wrote:
As per http://research.swtch.com/macpprof
We deliver all prof signals to the main thread, which is unlikely to
result in accurate profiling info. I think the diff below fixes things.
How about we take an idea from FreeBSD and have hardclock() just set
. IMO.
(I don't get why it's useful for tun0-in-layer3 mode to have the same
if_index as tun0-in-layer2 mode. The properties are so different that
there doesn't really seem to be continuity of identity between them.)
Philip Guenther
** yes, fixing ptrace() reparenting back to the original
On Fri, 16 Aug 2013, Ted Unangst wrote:
As per http://research.swtch.com/macpprof
We deliver all prof signals to the main thread, which is unlikely to
result in accurate profiling info. I think the diff below fixes things.
Making ITIMER_PROF per-thread like that matches neither what other
On Fri, Aug 16, 2013 at 12:23 AM, Ted Unangst t...@tedunangst.com wrote:
On Thu, Aug 15, 2013 at 23:39, Philip Guenther wrote:
Making ITIMER_PROF per-thread like that matches neither what other OS's
do
nor POSIX. jsing@ pinged about this last week and my comment then was
that this seems
On Tue, Aug 6, 2013 at 11:40 AM, Christopher Zimmermann
madro...@gmerlin.de wrote:
I'm currently working on the OpenBSD ip-stack for the first time.
Here's a small cosmetic improvement.
OK?
No, because the system doesn't build with that.
Philip Guenther
On Thu, Jul 4, 2013 at 2:54 AM, Artturi Alm artturi@gmail.com wrote:
old has been old and unused for some time already.
Committed; thanks!
Philip Guenther
until you rebuild with this diff.
oks?
Philip Guenther
Index: pattern.c
===
RCS file: /cvs/src/usr.bin/less/pattern.c,v
retrieving revision 1.1.1.1
diff -u -p -r1.1.1.1 pattern.c
--- pattern.c 16 Sep 2011 17:47:07 -
on the no onboard USB ports after resume problem, but
didn't expect any.
Philip Guenther
is wrong)
I can't figure out what you have in mind for this.
Philip Guenther
On Wed, Jul 17, 2013 at 6:10 PM, Alexander Bluhm
alexander.bl...@gmx.net wrote:
$ vmstat -M bsd.0.core -N bsd.0
vmstat: undefined symbols: _uvm_km_pages_free
There is no uvm_km_pages_free in our kernel. The corresponding
lines in dopool_sysctl() have been removed from vmstat.c revision
diff.
But here's a challenge for you: there's a case when your diff would be
better than mine, returning correct info that mine wouldn't. Do you
understand your own diff well enough to be able to what that case is?
Philip Guenther
;
if (sysctl(mib, 4, vfc, size, (void *)0, (size_t)0) 0) {
if (errno != EOPNOTSUPP)
That make sense?
Philip Guenther
On Sat, Jul 6, 2013 at 12:29 AM, Jason McIntyre j...@kerhand.co.uk wrote:
diff below from Daniel Dickman. it appears is* are functions, not macros
(defined in gen/isctype.c). i'm looking for oks.
Oh no! Someone found out that we changed them! We must hide the evidence!
Uh, I mean, sure,
); \
+ comp = (cmp)(elm, tmp); \
Your diff doesn't do what you say it does: adding parens would keep a
functional macro from being used.
Philip Guenther
On Sat, May 4, 2013 at 6:32 PM, Ted Unangst t...@tedunangst.com wrote:
We have never implemented amd64_get_ioperm and amd64_set_ioperm. There
are libarch stubs, but the kernel support has never been enabled. I'm
guessing nobody will miss it when it's gone.
ok guenther@
The man page also
On Sun, Apr 21, 2013 at 12:15 PM, Stefan Fritsch s...@sfritsch.de wrote:
On Sunday 21 April 2013, Mark Kettenis wrote:
...
You can't put ptrdiff_t in sys/types.h either.
Why not? POSIX only speaks about at least the following types in
sys/types.h, so there is no reason why it may not define
WILL NOT RUN YOUR INSTALLED
USERLAND. If that happens, you will have to boot from the old kernel
(do you know how?) or boot from an alternate device like a CD.
If you are not 100% comfortable with doing that recovery, then you
should wait and install the next snapshot instead.
Philip Guenther
On Sun, Apr 14, 2013 at 2:02 PM, Jérémie Courrèges-Anglas
j...@wxcvbn.org wrote:
Philip Guenther guent...@gmail.com writes:
...
As for -z now: yeah, it works on OpenBSD; what problem are they trying
to solve by using it?
-z now seems to be useful when used together with -z relro, IIUC
.
(There's one detail, how the address ranges are identified, that is
cleaner in the relro design. We're looking at pulling that bit into
how OpenBSD does it.)
As for -z now: yeah, it works on OpenBSD; what problem are they trying
to solve by using it?
Philip Guenther
(struct rtentry *, struct rttimer *);
+/* from ip_input.c */
+extern u_char ip_protox[];
+
extern struct protosw inetsw[];
No header file love for ip_protox and inetsw?
(Ideally, IMO, the end result would only have 'extern' in top-level
declarations in .h files)
Philip Guenther
On Mon, Apr 8, 2013 at 6:17 AM, Silamael silam...@coronamundi.de wrote:
I think the check whether the CPU has AES builtin misses an #ifdef
CRYPTO as it is there for the declaration of amd64_has_aesni.
Yep. Diff committed. Thanks!
Philip Guenther
. Diff committed. Thanks!
Philip Guenther
?
Philip Guenther
no program or script should be coded to rely on it.
Which reminds me: please update sysctl(3) and sysctl(8) too.
Philip Guenther
On Sun, Mar 31, 2013 at 11:49 PM, Jason McIntyre j...@kerhand.co.uk wrote:
pesky amino acids! ok?
...
The chemical name for tryptophan synthetase A protein, a
1,913-letter enzyme with 267 amino acids.
-- Mrs. Byrne's Dictionary of Unusual, Obscure, and
Hmm,
data structures after what routing sockets do on the wire.
Reusing the type of an double-underbar-prefix *padding* member gets you
hurt in back alleys.
Note that the stuff that it makes sense to copy into your own data
structures does use real, public types.
Philip Guenther
ok guenther@
The call in socreate() makes me think PRU_ATTACH should be moved from the
usrreq callback to a separate callback. Those casts... yuck...
On Fri, Mar 29, 2013 at 8:48 PM, Alexander Bluhm alexander.bl...@gmx.netwrote:
Hi,
Restrict protocol numbers for raw sockets to the range
.
Philip Guenther
this:
%l %{pie:-pie} %{p|pg|nopie:-nopie}
to this:
%{pie:-pie} %{p|pg|nopie:-nopie} %l
in the link_command spec.
...but I also like Mark's suggestion, to make -nopie not change the
link mode (shared-object vs executable).
Philip Guenther
On Thu, Mar 21, 2013 at 6:50 PM, David Hill dh...@mindcry.org wrote:
use %u for unsigned (ino_t).
Mumble mumble 64bit ino_t in the future mumble mumble
On Fri, Mar 22, 2013 at 7:53 AM, Gabriel Linder lin...@jeuxvideo.com wrote:
- * bpf understands a program iff kernel_major == filter_major
+ * bpf understands a program if kernel_major == filter_major
Not a typo: iff is a shorthand for if and only if that was
inherited from the mathematics
On Thu, Mar 14, 2013 at 9:20 PM, Jonathan Gray j...@jsg.id.au wrote:
The following diff changes wakeup so it will return
the number of processes it has woken up. Mostly
useful for debugging and error checking.
I initialy had a seperate function to check for a pending
wakeup but switched to
section, which says:
If s is a null pointer, the wcrtomb() function shall be equivalent
to the call:
wcrtomb(buf, L'\0', ps)
where buf is an internal buffer.
Since that call returns 1 when you pass in a buffer of your own,
wcrtomb(NULL, whatever, ps) should return 1 too.
Philip
The fix for this is (finally) committed for post-5.3. Sorry about the delay.
Philip Guenther
On Mon, 28 Jan 2013, Brad Smith wrote:
Here is a diff to add support for the POSIX functions dprintf(3)
and vdprintf(3).
Having poked at it a bit, I think the vdprintf.c below may be better. By
using its own write callback, it avoids the SHRT_MAX limit on the fd, and
it removes the fcntl()
I've had this exact diff, same whitespace and all, floating in my tree
since Aug 2011.
ok guenther@
. Adding an argument
(...and fixing the incorrect function header comment...) is the Right
Thing, IMO.
ok?
Philip Guenther
Index: display.c
===
RCS file: /cvs/src/usr.bin/top/display.c,v
retrieving revision 1.43
diff -u -p -r1.43
block for falsecell has 6 fields, but Cell has 7
fields: 2, 12, 0, 0, 0.0, 01
Patches to awk should be sent to the upstream maintainer listed in
/usr/src/usr.bin/awk/README
(I would tend to defer to that person's opinion on points of C style...)
Philip Guenther
On Tue, Dec 25, 2012 at 6:55 AM, Vadim Zhukov persg...@gmail.com wrote:
A small nit in fsck_ffs/dir.c. A few lines above (size_t) cast is
used, but here it's missing.
...
- if (memcmp(dirp-d_name, idesc-id_name, (int)dirp-d_namlen + 1))
+ if (memcmp(dirp-d_name, idesc-id_name,
* f_namemax. Looks like
FAT, for example, returns zero right now...
Philip Guenther
NAME_MAX path.)
Philip Guenther
On Fri, Nov 16, 2012 at 11:45 AM, Philip Guenther guent...@gmail.com wrote:
The diff below changes the alt sig stack logic to dynamically determine
whether the thread is currently on the alt stack, by comparing the stack
pointer against the altstack base and size, so that you get the correct
On Wed, Nov 28, 2012 at 1:42 PM, Mark Kettenis mark.kette...@xs4all.nl wrote:
But currently /dev/tunN is usable from any programming language that
that can do reads and writes. With Reyk's changes you need to do an
ioctl even for basic usage, which is at best quirky in languages other
than
On Fri, 23 Nov 2012, Alexander Hall wrote:
On 11/23/12 02:17, Philip Guenther wrote:
...
The argument to tolower() must be a value in the range [EOF,
0..UCHAR_MAX]. When taking characters from a char * string, you need
to cast the value to (unsigned char), ala
tolower((unsigned
tolower((unsigned char)*src)
Philip Guenther
to fix that, IMO.)
Philip Guenther
(), where ULONG_MAX is returned on error, but
can also be returned on success, meaning you have to do extra steps
(clear errno before, test it afterwards) to detect failure. This
sentence is the one that makes it clear that that isn't the case with
mmap().
Philip Guenther
On Thu, 27 Sep 2012, Alexey Suslikov wrote:
On Thursday, September 27, 2012, Alexey Suslikov wrote:
On Thursday, September 27, 2012, Philip Guenther wrote:
On Thu, 27 Sep 2012, Alexey Suslikov wrote:
Removing only local variables part reverts us to previous behavior
(i.e. crashes
On Wed, Sep 26, 2012 at 12:01 PM, Ted Unangst t...@tedunangst.com wrote:
I don't see what we gain by killing jobs. If the scheduler dice had
come down differently, maybe those jobs would finish.
Here's a downside, albeit maybe a stretch. What if the job doesn't
like being killed? You're
the crash? Where did the bogus pointer come from?
I'm starting to believe that static globals are not good.
They are incredibly good at what they do. If you're trying to say that
they fundamentally can't be thread-safe, you'll need some extraordinary
evidence for such a claim.
Philip Guenther
the documented method of getting the defines for the sysctl
nodes under CTL_HW: c.f., both sysctl(3) and sysctl(8). That actually
seems appropriate for this.
Philip Guenther
On Tue, 18 Sep 2012, Marc Espie wrote:
On Tue, Sep 18, 2012 at 01:39:14PM -0600, Theo de Raadt wrote:
Standards gods are capricious gods.
They are false gods.
Sounds like American gods :)
Maybe Standards is the god that no one can remember that Wednesday meets
in Las Vegas...
On Sun, Sep 16, 2012 at 9:18 AM, Jason McIntyre j...@kerhand.co.uk wrote:
just spotted that the name description (Nd) for this file is now out of
date. updated accordingly (from freebsd).
ok?
Ship it!
guenther@
On Sat, Sep 15, 2012 at 1:42 AM, Jason McIntyre j...@kerhand.co.uk wrote:
On Sat, Sep 15, 2012 at 10:19:35AM +0200, Michal Mazurek wrote:
doesn;t the code mean the check is for zero, not less than or equal to
zero?
Oops, you're right.
everything else looks alright. i'll commit it soon if
be
compiled, the resulting depend file will not list this source. Without
having any dependencies, make(1) will not try to rebuild the derived
targets. Unless I do a 'make clean', I won't even notice that my
build is broken.
Below is a possible patch to fix this.
Committed; thanks!
Philip Guenther
that includes
that test. :-)
I think the lex.h diff is nicer if the new SBRACEQ state is stuck at
the end of the existing list instead of renumbering additional states,
but I can be argued into renumbering them...
Philip Guenther
the thread's current stack pointer, which
makes the handling of the alt-stack more flexible without creating any
contradictions. We should pull that change over, IMO.
Philip Guenther
On Sun, 12 Aug 2012, Alexander Polakov wrote:
Nobody cares, so here's my attempt. I haven't run it though a
world/ports build because of slow hardware on my hands right now.
Index: lex.c
===
RCS file: /cvs/src/bin/ksh/lex.c,v
. The bug applies to the simple double-quoted case too:
$ foo=1
$ echo ${foo:+'blah'}
blah
$
The correct output there should include the single-quotes, ala:
'blah'
Philip Guenther
?
hahahhahahhahahahahahh
Philip Guenther
the types...
Philip Guenther
about the memory usage of the system.
Philip Guenther
On Sat, Jun 23, 2012 at 10:43 PM, Seth Wright s...@crosse.org wrote:
Looks like someone wasn't sure which verb to use, so they went ahead
and used both.
Committed. Thanks!
Philip Guenther
On Fri, Jun 22, 2012 at 2:41 AM, Paul Irofti p...@irofti.net wrote:
Tested with the following program and everything seems to work. I've
also sprinkled some extra printfs in the original diff and both cases
are found and dealt with in the kernel.
Woot. In it goes!
that, then it's better
than what's there.
Philip Guenther
and SOCK_NONBLOCK? I would expect ignoring
the former to result in rapid fd leaks, and ignoring of the latter to
result in code blocking unexpectedly...
Philip Guenther
.
Peter, do you want to take a stab at that part too?
(There's also a typo in the current page: s/HOPTLIMIT/HOPLIMIT/)
Philip Guenther
will
probably better match expectations, however.
Philip Guenther
On Wed, 11 Apr 2012, Mike Belopuhov wrote:
guenther@ has pointed out that sometimes suser is passed an incorrect
flags argument. Currently, there's only one flag that's defined:
SUSER_NOACCT (0x1). Unfortunately it aligns too well with AFORK flag
used with p_acflag. This diff cleans up
hardware.
Philip Guenther
On Mon, Mar 12, 2012 at 8:45 PM, Ted Unangst t...@tedunangst.com wrote:
run fstat -u root. observe non root files.
I don't know why you would pass -1 here to get all files when there's
better ways to do that, but just the same, I won't change that.
Index: kern_sysctl.c
patch that makes it abort on the first error is
certainly not acceptable.
Philip Guenther
Now that the base is unlocked, we can consider diffs to
a) rename getdelim() and getline() symbols that are already in the tree, and
b) add getdelim() and getline() to libc.
Note that (a) will need to go in before (b)...
Philip Guenther
On Thu, Feb 23, 2012 at 5:56 PM, Tobias Ulmer tobi...@tmux.org wrote:
Anyone brave enough to commit this diff or should I put it into the
round file? :)
Committed. Thanks for the poke...
Philip Guenther
for the upcoming 5.1
release. Thank you for reporting it!
Philip Guenther
identifiers are broken.
blambert@ should be committing the fix soon.
Philip Guenther
is reached and that programs that want
to tell apart end-of-directory and errors must set errno to zero
should be added.
I also think readdir() should set errno if it detects an invalid
seekdir(). EINVAL seems correct.
Philip Guenther
On Thu, 2 Feb 2012, Philip Guenther wrote:
I also think readdir() should set errno if it detects an invalid
seekdir(). EINVAL seems correct.
Here's a diff for this bit.
oks?
Philip Guenther
Index: gen/readdir.c
===
RCS file
On Thu, Jan 19, 2012 at 4:18 AM, Edd Barrett vex...@gmail.com wrote:
I recently got caught out by scandir's quirky memory allocation. Should
we add a note so that others don't have to go through this pain too?
Could you elaborate on how this affected your code? I'm trying to
think of some way
that it should always show itself to
poll() as writable. Now, it may be that the packet will be thrown
away if the output queue is backed up, but that's not the question
that poll() asks generally, so that's fine.
Philip Guenther
://pubs.opengroup.org/onlinepubs/9699919799/functions/pthread_mutex_timedlock.html
)
Philip Guenther
that gets it and, IMO, it's clearer for the
message to be sent to stdout with the rest of the output from the
scan.
As for errors other than EPERM, well, exactly what other errors *can*
that call return in ifconfig?
What problem are you guys trying to solve?
Philip Guenther
$ ifconfig rsu0 scan
rsu0
to do the reset mon to zero and bump
year dance. Just
if (l tm.tm_mday)
tm.tm_mon++;
is sufficient.
Philip Guenther
H
+Also display information about kernel visible threads.
.It Fl h
Repeat the information header as often as necessary to guarantee one
header per page of information.
Philip Guenther
*ent.c code). Thanks!
Philip Guenther
This diff adds to touch(1) support for the -d option specified by POSIX
2008 that permits specifying subsecond timestamps, as well as nanosecond
support in the -r option, and a general simplification over the main loop.
oks?
Philip
Index: usr.bin/touch/touch.1
with /usr/src on it is dieing, then the first step would be to take a
backup and install a replacement drive; if it was just the result of
mounting /usr/obj async and not cleaning up after a crash then a
simple 'make cleandir' would be enough.
Philip Guenther
...)
Philip Guenther
On Mon, Jul 4, 2011 at 12:50 PM, Owain Ainsworth zer...@googlemail.com
wrote:
So, the msync implementation in uvm says this:
/*
* XXXCDC: do we really need this semantic?
*
* XXX Gak! If size is zero we are supposed to sync all modified
* pages with the
header (say, _null.h) and then all those
files would include that instead.
Philip Guenther
In libpthread, if the scheduler picks the current thread as the next to
run, it still does the full save+restore of errno, FPU, and regs, saving
them and then restoring the same values again. That's kinda silly, so
jump around that if switching to the current thread.
oks?
Philip Guenther
First, expand the INTR_RESTORE_GPRS macro and then rearrange the register
restores around the INTR_RESTORE_SELECTORS macro to minimize how long
interupts are blocked. This also lets us eliminate all the adjusting of
the stack pointer except for the necessary one on the iretq path.
Instead of
with a SIGKILL between the fds, right?
Philip Guenther
On Thu, Jun 2, 2011 at 8:52 AM, Paul Irofti p...@irofti.net wrote:
Any objections to adding this to the LIST structure?
Yes, because it doesn't work.
If you want that, just change it from a LIST to a TAILQ.
Philip Guenther
On Thu, May 26, 2011 at 2:17 AM, Nicholas Marriott
nicholas.marri...@gmail.com wrote:
kq attach functions should return an errno so 1 is wrong
spotted while looking at kq/pipe problem, there are lots of others in
the tree but that's for later
ok?
Yeah, seems obviously correct to me. ok
or
its a relic?
Neither.
i was looking for the equivalent of FreeBSD's file of the same name and
location but in OpenBSD.
Our file serves the same purpose as FreeBSD's, it's just incomplete.
What do you think FreeBSD is for/does?
Philip Guenther
to
between the addl and the iret would be enough. If you can reproduce
this situation easily than I would ask that you try that.
Possibility two would be to try to handle this from, uh, inside
resume_iret I guess. I'm not 100% sure whether it can be unilateral
there though.
Philip Guenther
=$LC_ALL
A quick grep of /etc/*/* on a Redhat box finds *lots* of LANG=C
settings on commands. Wheee
Philip Guenther
will reveal something about a recent commit.
...
Stopped at fill_kproc+0x24e: movl 0x19c(%rax),%eax
That was my botch: missed a case while reworking the signal handling
for correctness with rthreads. It's already fixed in the tree and the
next snapshot will surely include the fix.
Philip
801 - 900 of 983 matches
Mail list logo