Re: [PATCH] handle return value of clock_nanosleep(2) correctly

  • From: Hajime Tazaki<tazaki@xxxxxxxxxx>
  • To: pooka@xxxxxx
  • Date: Thu, 08 Oct 2015 17:20:51 +0900


At Thu, 8 Oct 2015 06:44:59 +0000,
Antti Kantee wrote:


On 08/10/15 05:35, tazaki@xxxxxxxxxx wrote:
From: Hajime Tazaki <thehajime@xxxxxxxxx>

when nanosleep(2) is used in rumpuser_clock_sleep(), it will properly
retry if errno == EINTR, but with clock_nanosleep(2) it won't. This
patch aligns the error values of clock_nanosleep() to nanosleep's case.

Thanks, but unfortunately(?) this bug has already been fixed in the
NetBSD tree. When sending patches against the POSIX rumpuser
implementation, please send them against the NetBSD HEAD.

Ah, thanks. Good news.
I will check NetBSD HEAD next time.


btw, clock_nanosleep(2) seems to be a bit trickier.

NetBSD returns -1 while Linux returns errno (positive).

http://man.netbsd.org/HEAD/usr/share/man/html2/nanosleep.html
http://man7.org/linux/man-pages/man2/clock_nanosleep.2.html

need an #ifdef around the code ? shiver...

-- Hajime


Other related posts:

  1. Home
  2. rumpkernel-users
  3. Archive
  4. 10-2015