public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Saurabh Desai <sdesai@austin.ibm.com>
To: linux-kernel@vger.kernel.org, mingo@elte.hu
Cc: phil-list@redhat.com
Subject: [PATCH]fix for thread signal delivery
Date: Fri, 31 Jan 2003 11:37:17 -0600	[thread overview]
Message-ID: <3E3AB44D.D4B9DC22@austin.ibm.com> (raw)

[-- Attachment #1: Type: text/plain, Size: 510 bytes --]


A signal delivery to a thread is not working
properly in the kernel when a signal 
handler is specified. Here is the case:
A process blocks all signals and registers
a handler for a signal. 
A newly created thread unblocks that signal.
When that signal is sent to a process, the 
sighandler is not getting called.
(if this signal is sent twice, it works)
  
I've attached a small test program to reproduce 
it using NPTL. 
The following patch fixes this problem in kernel.
Please apply.

Thanks,
Saurabh Desai

[-- Attachment #2: sigfix.diff --]
[-- Type: text/plain, Size: 543 bytes --]

diff -Naur linux-2.5.59.orig/kernel/signal.c linux-2.5.59/kernel/signal.c
--- linux-2.5.59.orig/kernel/signal.c	2003-01-30 16:54:44.000000000 -0600
+++ linux-2.5.59/kernel/signal.c	2003-01-30 16:55:22.000000000 -0600
@@ -854,11 +854,13 @@
 
 	tmp = p->sig->curr_target;
 
-	if (!tmp || tmp->tgid != p->tgid)
+	if (!tmp || tmp->tgid != p->tgid) {
 		/* restart balancing at this thread */
 		p->sig->curr_target = p;
+		tmp = p;
+	}
 
-	else for (;;) {
+	for (;;) {
 		if (thread_group_empty(p))
 			BUG();
 		if (!tmp || tmp->tgid != p->tgid)

[-- Attachment #3: tsig.c --]
[-- Type: text/plain, Size: 1036 bytes --]


#include <stdlib.h>
#include <stdio.h>
#include <unistd.h>
#include <pthread.h>
#include <signal.h>

void sigcatcher(int sig)
{
    fprintf(stderr, "The sigcatcher caught signal %d\n", sig);
    pthread_exit(0);
}

void *sig_thread(void *p)
{
    sigset_t sigs_to_catch;

    /* Unblock SIGINT */
    sigemptyset(&sigs_to_catch);
    sigaddset(&sigs_to_catch, SIGINT);
    pthread_sigmask(SIG_UNBLOCK, &sigs_to_catch, NULL);

    fprintf(stderr, "sent a SIGINT (ctrl-C) to this process(%d)\n", getpid());
    sleep(60);

    return (NULL);
}

extern int main(void)
{
    pthread_t thread;
    sigset_t sigs_to_block;
    struct sigaction action;

    /* Block all signals */
    sigfillset(&sigs_to_block);
    pthread_sigmask(SIG_BLOCK, &sigs_to_block, NULL);
    /* Set a signal handler */
    action.sa_handler = sigcatcher;
    action.sa_flags = 0;
    sigaction(SIGINT, &action, NULL);

    pthread_create(&thread, NULL, sig_thread, NULL);

    /* wait until thread is finished */
    pthread_join(thread, NULL);

    return 0;
}

                 reply	other threads:[~2003-01-31 17:28 UTC|newest]

Thread overview: [no followups] expand[flat|nested]  mbox.gz  Atom feed

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=3E3AB44D.D4B9DC22@austin.ibm.com \
    --to=sdesai@austin.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=phil-list@redhat.com \
    /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