qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Avi Kivity <avi@redhat.com>
To: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [patch 1/7] qemu: mutex/thread/cond wrappers
Date: Thu, 19 Mar 2009 17:59:37 +0200	[thread overview]
Message-ID: <49C26BE9.4030905@redhat.com> (raw)
In-Reply-To: <20090319150537.801001000@localhost.localdomain>

mtosatti@redhat.com wrote:
> --- qemu.orig/vl.c
> +++ qemu/vl.c
> @@ -36,6 +36,7 @@
>  #include "gdbstub.h"
>  #include "qemu-timer.h"
>  #include "qemu-char.h"
> +#include "qemu-thread.h"
>  #include "cache-utils.h"
>  #include "block.h"
>  #include "audio/audio.h"
> @@ -263,6 +264,8 @@ static QEMUTimer *nographic_timer;
>  
>  uint8_t qemu_uuid[16];
>  
> +QemuMutex qemu_global_mutex;
> +
>  /***********************************************************/
>  /* x86 ISA bus support */
>  
> @@ -3650,7 +3653,14 @@ void main_loop_wait(int timeout)
>          slirp_select_fill(&nfds, &rfds, &wfds, &xfds);
>      }
>  #endif
> +
> +    /*
> +     * main_loop_wait() *must* not assume any global state is consistent across
> +     * select() invocations.
> +     */
> +    qemu_mutex_unlock(&qemu_global_mutex);
>      ret = select(nfds + 1, &rfds, &wfds, &xfds, &tv);
> +    qemu_mutex_lock(&qemu_global_mutex);
>      if (ret > 0) {
>          IOHandlerRecord **pioh;
>  
> @@ -3708,6 +3718,9 @@ static int main_loop(void)
>  #endif
>      CPUState *env;
>  
> +    qemu_mutex_init(&qemu_global_mutex);
> +    qemu_mutex_lock(&qemu_global_mutex);
> +
>      cur_cpu = first_cpu;
>      next_cpu = cur_cpu->next_cpu ?: first_cpu;
>      for(;;) {
>   

All the bits above seem unnecessary for the this patch.


> +
> +int qemu_mutex_init(QemuMutex *mutex)
> +{
> +    return pthread_mutex_init(&mutex->lock, NULL);
> +}
>   

No one will check the return code, so better to check it here and 
abort() if it's bad.

> +static void add_to_timespec(struct timespec *ts, unsigned int msecs)
> +{
> +    ts->tv_sec = ts->tv_sec + (long)(msecs / 1000);
> +    ts->tv_nsec = (ts->tv_nsec + ((long)msecs % 1000) * 1000000);
> +    if (ts->tv_nsec >= 1000000000) {
> +        ts->tv_nsec -= 1000000000;
> +        ts->tv_sec++;
> +    }
> +}
>   

timespec_add_msec()?  and why milliseconds?  Also, maybe uint64_t is a 
better type.

> +
> +int qemu_mutex_timedlock(QemuMutex *mutex, unsigned int msecs)
> +{
> +    struct timespec ts;
> +
> +    clock_gettime(CLOCK_REALTIME, &ts);
> +    add_to_timespec(&ts, msecs);
> +
> +    return pthread_mutex_timedlock(&mutex->lock, &ts);
> +}
>   

I would have preferred a deadline instead of a timeout, but we'll see on 
the next patches.

> +
> +int qemu_thread_create(QemuThread *thread,
> +                       void *(*start_routine)(void*),
> +                       void *arg)
> +{
> +    return pthread_create(&thread->thread, NULL, start_routine, arg);
> +}
>   

Don't return an error that no one will check.

> +
> +int qemu_thread_signal(QemuThread *thread, int sig)
> +{
> +    if (thread->thread != 0)
> +        return pthread_kill(thread->thread, sig);
> +    return -1; /* XXX: ESCHR */
> +}
>   

Ditto.  If the thread dies, qemu dies.

> +int qemu_thread_equal(QemuThread *thread1, QemuThread *thread2)
> +{
> +   return (thread1->thread == thread2->thread);
> +}
>   

Can compare thing1 to thing2 instead of ->thread.  Not that it matters.


-- 
error compiling committee.c: too many arguments to function

  reply	other threads:[~2009-03-19 15:59 UTC|newest]

Thread overview: 19+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-03-19 14:57 [Qemu-devel] [patch 0/7] separate thread for io v2 mtosatti
2009-03-19 14:57 ` [Qemu-devel] [patch 1/7] qemu: mutex/thread/cond wrappers mtosatti
2009-03-19 15:59   ` Avi Kivity [this message]
2009-03-19 18:49     ` Jamie Lokier
2009-03-23 23:17     ` Marcelo Tosatti
2009-03-24  7:43       ` Avi Kivity
2009-03-19 14:57 ` [Qemu-devel] [patch 2/7] qemu: separate thread for io mtosatti
2009-03-20 17:43   ` [Qemu-devel] " Anthony Liguori
2009-03-21  0:06     ` Marcelo Tosatti
2009-03-21  1:04       ` Anthony Liguori
2009-03-21  1:44         ` Marcelo Tosatti
2009-03-21  1:50           ` Anthony Liguori
2009-03-22  8:48             ` Avi Kivity
2009-03-22 11:17               ` Anthony Liguori
2009-03-19 14:57 ` [Qemu-devel] [patch 3/7] qemu: main thread does io and cpu thread is spawned mtosatti
2009-03-19 14:57 ` [Qemu-devel] [patch 4/7] qemu: handle reset/poweroff/shutdown in iothread mtosatti
2009-03-19 14:57 ` [Qemu-devel] [patch 5/7] qemu: pause and resume cpu thread(s) mtosatti
2009-03-19 14:57 ` [Qemu-devel] [patch 6/7] qemu: handle vmstop from cpu context mtosatti
2009-03-19 14:57 ` [Qemu-devel] [patch 7/7] qemu: use pipe to wakeup io thread mtosatti

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=49C26BE9.4030905@redhat.com \
    --to=avi@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).