Re: ihidev: always register interrupt, stop polling if it fires

2019-07-19 Thread Mike Larkin
On Fri, Jul 19, 2019 at 02:16:18PM -0500, joshua stein wrote: > The fast polling of ihidev may cause a problem during suspend/resume > because dwiic may be in an unknown state, so add a DVACT_QUIESCE > handler to properly shut it down. > > Even if polling is requested, register the interrupt

Re: [patch] Re-add 'proc' to vi(1) when running in secure mode.

2019-07-19 Thread Bryan Steele
On Sat, Jul 20, 2019 at 12:03:03AM +0200, Jesper Wallin wrote: > On Fri, Jul 19, 2019 at 05:14:03PM -0400, Bryan Steele wrote: > > I suspect that in secure/-S mode, the :pre[serve] should either be > > disabled, or modified to stop calling sendmail. The mail it is sending > > is purely advisory,

Re: [patch] Re-add 'proc' to vi(1) when running in secure mode.

2019-07-19 Thread Jesper Wallin
On Fri, Jul 19, 2019 at 05:14:03PM -0400, Bryan Steele wrote: > I suspect that in secure/-S mode, the :pre[serve] should either be > disabled, or modified to stop calling sendmail. The mail it is sending > is purely advisory, and should be easy to disable. See common/recover.c. Oh, you're right.

Re: [patch] Re-add 'proc' to vi(1) when running in secure mode.

2019-07-19 Thread Bryan Steele
On Fri, Jul 19, 2019 at 09:43:14PM +0200, Jesper Wallin wrote: > Hi all, > > When using vi(1) with secure mode (-S), both 'proc' and 'exec' are > stripped from the pledge promise. This breaks the :pre[serve] command > as it uses fork(2). This is broken on 6.4, 6.5 and -current. > > Re-add the

[patch] Re-add 'proc' to vi(1) when running in secure mode.

2019-07-19 Thread Jesper Wallin
Hi all, When using vi(1) with secure mode (-S), both 'proc' and 'exec' are stripped from the pledge promise. This breaks the :pre[serve] command as it uses fork(2). This is broken on 6.4, 6.5 and -current. Re-add the 'proc' promise, even when running in secure mode. Jesper Wallin Index:

ihidev: always register interrupt, stop polling if it fires

2019-07-19 Thread joshua stein
The fast polling of ihidev may cause a problem during suspend/resume because dwiic may be in an unknown state, so add a DVACT_QUIESCE handler to properly shut it down. Even if polling is requested, register the interrupt handler too. If it ever fires, stop polling and just use the interrupt.

Re: ping(8): don't round timeouts up to HZ

2019-07-19 Thread Todd C . Miller
On Fri, 19 Jul 2019 10:29:11 -0500, Scott Cheloha wrote: > ping(8) doesn't know the value of HZ and shouldn't round the timeout > up. The kernel will do the right thing anyway. > > Not sure if the warning here is useful either. If the period is too > small it will just fire less often. We

ping(8): don't round timeouts up to HZ

2019-07-19 Thread Scott Cheloha
ping(8) doesn't know the value of HZ and shouldn't round the timeout up. The kernel will do the right thing anyway. Not sure if the warning here is useful either. If the period is too small it will just fire less often. We could add a warning about it in CAVEATS in ping.8... Preferences?

Re: systat is not flushing output

2019-07-19 Thread Todd C . Miller
On Fri, 19 Jul 2019 17:16:00 +0200, Solene Rapenne wrote: > I wanted to use systat -b -d 10 iostat 1 | something to parse output in > real time but it was not possible, systat doesn't flush output. > > otto@ suggested this change, I tried it and it works. > > This flush output at the end of

systat is not flushing output

2019-07-19 Thread Solene Rapenne
I wanted to use systat -b -d 10 iostat 1 | something to parse output in real time but it was not possible, systat doesn't flush output. otto@ suggested this change, I tried it and it works. This flush output at the end of writing a page in raw mode (-B or -b flags). ok? Index: engine.c

Re: fix: NULL dereference in bios(4)

2019-07-19 Thread Jan Klemkow
On Fri, Jul 19, 2019 at 09:13:38PM +1000, Jonathan Gray wrote: > On Fri, Jul 19, 2019 at 01:07:34PM +0200, Jan Klemkow wrote: > > Hi, > > > > fixstring() can return NULL and it does on one of my machines. > > Here is s fix that checks for NULL and return an empty string. > > > > ok? > > If it

Re: fix: NULL dereference in bios(4)

2019-07-19 Thread Jonathan Gray
On Fri, Jul 19, 2019 at 01:07:34PM +0200, Jan Klemkow wrote: > Hi, > > fixstring() can return NULL and it does on one of my machines. > Here is s fix that checks for NULL and return an empty string. > > ok? If it returns NULL shouldn't the strlcpy and printf both be skipped? You've also not

fix: NULL dereference in bios(4)

2019-07-19 Thread Jan Klemkow
Hi, fixstring() can return NULL and it does on one of my machines. Here is s fix that checks for NULL and return an empty string. ok? bye, Jan Index: arch/amd64/amd64/bios.c === RCS file: /cvs/src/sys/arch/amd64/amd64/bios.c,v

Re: taking kernel config into consideration when reorder

2019-07-19 Thread Gregory Edigarov
On 19.07.19 11:12, Gregory Edigarov wrote: On 18.07.19 19:38, Solene Rapenne wrote: On Thu, Jul 18, 2019 at 07:15:50PM +0300, Gregory Edigarov wrote: Hello, Just tired a of rebooting into UKC a bit. diff --git a/libexec/reorder_kernel/reorder_kernel.sh

Re: taking kernel config into consideration when reorder

2019-07-19 Thread Gregory Edigarov
On 18.07.19 19:38, Solene Rapenne wrote: On Thu, Jul 18, 2019 at 07:15:50PM +0300, Gregory Edigarov wrote: Hello, Just tired a of rebooting into UKC a bit. diff --git a/libexec/reorder_kernel/reorder_kernel.sh b/libexec/reorder_kernel/reorder_kernel.sh index ecd8d8fc563..7354350505a 100644