From: Davide Libenzi <davidel@xmailserver.org>
To: Linus Torvalds <torvalds@osdl.org>
Cc: Roland McGrath <roland@redhat.com>,
Andreas Schwab <schwab@suse.de>,
Daniel Jacobowitz <dan@debian.org>,
Eric Pouech <pouech-eric@wanadoo.fr>,
Mike Hearn <mh@codeweavers.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Andrew Morton <akpm@osdl.org>, wine-devel <wine-devel@winehq.com>
Subject: Re: ptrace single-stepping change breaks Wine
Date: Mon, 22 Nov 2004 05:46:42 -0800 (PST) [thread overview]
Message-ID: <Pine.LNX.4.58.0411220539270.19254@bigblue.dev.mdolabs.com> (raw)
In-Reply-To: <Pine.LNX.4.58.0411212212530.20993@ppc970.osdl.org>
On Sun, 21 Nov 2004, Linus Torvalds wrote:
> Ok, how about this patch?
>
> It does basically two things:
>
> - it makes the x86 version of ptrace be a lot more careful about the TF
> bit in eflags, and in particular it never touches it _unless_ the
> tracer has explicitly asked for it (ie we set TF only when doing a
> PTRACE_SINGESTEP, and we clear it only when it has been set by us, not
> if it has been set by the program itself).
>
> This patch also cleans up the codepaths by doing all the common stuff
> in set_singlestep()/clear_singlestep().
>
> - It clarifies signal handling, and makes it clear that we always push
> the full eflags onto the signal stack, _except_ if the TF bit was set
> by an external ptrace user, in which case we hide it so that the tracee
> doesn't see it when it looks at its stack contents.
>
> It also adds a few comments, and makes it clear that the signal handler
> itself is always set up with TF _clear_. But if we were single-stepped
> into it, we will have notified the debugger, so the debugger obviously
> can (and often will) decide to continue single-stepping.
Looks like a nice cleanup. What does the test program below print for you?
- Davide
#include <stdio.h>
#include <stdlib.h>
#include <signal.h>
#include <unistd.h>
#include <errno.h>
#include <sys/types.h>
#include <sys/ptrace.h>
#include <sys/wait.h>
#include <linux/user.h>
#include <linux/unistd.h>
int main(int ac, char **av) {
int i, status, res;
long start, end;
pid_t cpid, pid;
struct user_regs_struct ur;
struct sigaction sa;
sigemptyset(&sa.sa_mask);
sa.sa_flags = 0;
sa.sa_handler = SIG_DFL;
sigaction(SIGCHLD, &sa, NULL);
printf("nargs=%d\n", ac);
if (ac == 1)
goto tracer;
printf("arg=%s\n", av[1]);
loop:
__asm__ volatile ("int $0x80"
: "=a" (res)
: "0" (__NR_getpid));
goto loop;
endloop:
exit(0);
tracer:
if ((cpid = fork()) != 0)
goto parent;
printf("child=%d\n", getpid());
ptrace(PTRACE_TRACEME, 0, NULL, NULL);
execl(av[0], av[0], "child", NULL);
exit(0);
parent:
start = (long) &&loop;
end = (long) &&endloop;
printf("pchild=%d\n", cpid);
for (;;) {
pid = wait(&status);
if (pid != cpid)
continue;
res = WSTOPSIG(status);
if (ptrace(PTRACE_GETREGS, pid, NULL, &ur)) {
printf("[%d] error: ptrace(PTRACE_GETREGS, %d)\n",
pid, pid);
return 1;
}
if (ptrace(PTRACE_SINGLESTEP, pid, NULL, res != SIGTRAP ? res: 0)) {
perror("ptrace(PTRACE_SINGLESTEP)");
return 1;
}
if (ur.eip >= start && ur.eip <= end)
break;
}
for (i = 0; i < 15; i++) {
printf("waiting ...\n");
pid = wait(&status);
printf("done: pid=%d status=%d\n", pid, status);
if (pid != cpid)
continue;
res = WSTOPSIG(status);
printf("sig=%d\n", res);
if (ptrace(PTRACE_GETREGS, pid, NULL, &ur)) {
printf("[%d] error: ptrace(PTRACE_GETREGS, %d)\n",
pid, pid);
return 1;
}
printf("EIP=0x%08x\n", ur.eip);
if (ptrace(PTRACE_SINGLESTEP, pid, NULL, res != SIGTRAP ? res: 0)) {
perror("ptrace(PTRACE_SINGLESTEP)");
return 1;
}
}
if (ptrace(PTRACE_CONT, cpid, NULL, SIGKILL)) {
perror("ptrace(PTRACE_SINGLESTEP)");
return 1;
}
return 0;
}
next prev parent reply other threads:[~2004-11-22 13:50 UTC|newest]
Thread overview: 101+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <Pine.LNX.4.58.0411151439270.2222@ppc970.osdl.org>
2004-11-15 22:53 ` ptrace single-stepping change breaks Wine Roland McGrath
2004-11-19 19:00 ` Eric Pouech
2004-11-19 19:20 ` Linus Torvalds
2004-11-19 19:33 ` Eric Pouech
2004-11-19 19:51 ` Linus Torvalds
2004-11-19 20:41 ` Eric Pouech
2004-11-19 21:22 ` Linus Torvalds
2004-11-19 21:23 ` Daniel Jacobowitz
2004-11-19 21:53 ` Linus Torvalds
2004-11-20 21:49 ` Jesse Allen
2004-11-21 4:55 ` Jesse Allen
2004-11-21 21:32 ` Davide Libenzi
2004-11-21 22:33 ` Linus Torvalds
2004-11-21 23:14 ` Davide Libenzi
2004-11-22 1:12 ` Linus Torvalds
2004-11-22 0:13 ` Andreas Schwab
2004-11-22 1:07 ` Linus Torvalds
2004-11-22 4:06 ` Davide Libenzi
2004-11-22 4:29 ` Linus Torvalds
2004-11-22 6:23 ` Linus Torvalds
2004-11-22 11:06 ` Andreas Schwab
2004-11-22 16:27 ` Linus Torvalds
2004-11-22 13:46 ` Davide Libenzi [this message]
2004-11-22 23:15 ` Jesse Allen
2004-11-22 23:48 ` Jesse Allen
2004-11-28 17:01 ` Eric Pouech
2004-11-22 20:52 ` Eric Pouech
2004-11-22 21:10 ` Linus Torvalds
2004-11-22 22:19 ` Mike Hearn
2004-11-22 22:25 ` Linus Torvalds
2004-12-29 2:14 ` Thomas Sailer
2004-12-29 15:02 ` Mike Hearn
2004-12-29 18:53 ` Linus Torvalds
2004-12-29 19:40 ` Jesse Allen
2004-12-29 20:04 ` Linus Torvalds
2004-12-29 21:43 ` Jesse Allen
2004-12-30 0:44 ` Linus Torvalds
2004-12-30 1:13 ` Davide Libenzi
2004-12-30 1:55 ` Linus Torvalds
2004-12-30 4:51 ` Linus Torvalds
2004-12-30 4:58 ` Linus Torvalds
2004-12-30 5:07 ` Davide Libenzi
2004-12-30 7:26 ` Linus Torvalds
2004-12-30 17:59 ` Davide Libenzi
2004-12-30 18:16 ` Linus Torvalds
2004-12-30 19:27 ` Jesse Allen
2004-12-30 19:34 ` Linus Torvalds
2004-12-30 22:46 ` Linus Torvalds
2004-12-30 23:00 ` Daniel Jacobowitz
2004-12-30 23:17 ` Linus Torvalds
2004-12-31 5:36 ` Daniel Jacobowitz
2004-12-31 5:47 ` Linus Torvalds
2004-12-31 7:00 ` Jesse Allen
2004-12-31 15:10 ` Daniel Jacobowitz
2004-12-31 17:19 ` Linus Torvalds
2005-01-01 23:20 ` Daniel Jacobowitz
2005-01-29 9:25 ` Kari Hurtta
2004-12-30 23:15 ` Andi Kleen
2004-12-31 0:38 ` Linus Torvalds
2004-12-31 12:35 ` Andi Kleen
2004-12-31 15:16 ` Davide Libenzi
2004-12-31 17:30 ` Linus Torvalds
2004-12-31 19:55 ` Jesse Allen
2004-12-31 17:14 ` Linus Torvalds
2004-12-31 4:55 ` Jesse Allen
2004-12-31 5:05 ` Linus Torvalds
2004-12-31 5:38 ` Daniel Jacobowitz
2004-12-30 19:19 ` Davide Libenzi
2004-12-30 5:06 ` Davide Libenzi
2004-12-30 4:28 ` Jesse Allen
2004-12-29 20:56 ` Jesse Allen
2004-12-29 19:35 ` Thomas Sailer
2004-12-29 20:13 ` Jesse Allen
2004-12-30 1:49 ` Thomas Sailer
2004-12-30 2:10 ` Linus Torvalds
2004-12-30 2:39 ` Thomas Sailer
2004-12-30 2:57 ` Thomas Sailer
2004-12-30 3:15 ` Thomas Sailer
2004-12-30 4:15 ` Andrew Morton
2004-12-30 10:09 ` Thomas Sailer
2004-12-30 13:06 ` Mike Hearn
2004-12-31 13:13 ` Thomas Sailer
2004-12-31 13:31 ` Mike Hearn
2004-12-31 15:42 ` Jesse Allen
2004-12-31 15:56 ` Davide Libenzi
2004-12-31 15:59 ` Jesse Allen
2004-12-31 22:01 ` Linus Torvalds
2005-01-01 22:04 ` Davide Libenzi
2005-01-01 22:14 ` Linus Torvalds
2005-01-02 3:46 ` Davide Libenzi
2005-01-07 4:51 ` minor nit with decoding popf instruction - was " John Kacur
2005-01-07 6:48 ` Linus Torvalds
2005-01-08 5:05 ` John Kacur
2004-12-31 15:51 ` Thomas Sailer
[not found] ` <1104873315.3557.87.camel@littlegreen>
2005-01-04 21:21 ` Andrew Morton
2005-01-05 10:43 ` Thomas Sailer
2005-01-05 11:24 ` Ingo Molnar
2005-01-05 11:40 ` Alexandre Julliard
2004-12-30 12:11 ` Mike Hearn
2004-11-20 3:40 ` Roland McGrath
2004-11-19 20:59 ` Grzegorz Kulewski
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=Pine.LNX.4.58.0411220539270.19254@bigblue.dev.mdolabs.com \
--to=davidel@xmailserver.org \
--cc=akpm@osdl.org \
--cc=dan@debian.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mh@codeweavers.com \
--cc=pouech-eric@wanadoo.fr \
--cc=roland@redhat.com \
--cc=schwab@suse.de \
--cc=torvalds@osdl.org \
--cc=wine-devel@winehq.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;
as well as URLs for NNTP newsgroup(s).