DMARC-Filter: OpenDMARC Filter v1.4.2 delorie.com 52PDfFCV2668983
DMARC-Filter: OpenDMARC Filter v1.4.2 sourceware.org B3BD63858D33
Subject: Hang or crash after multiple SIGILL or SIGSEGV and siglongjmp
To: cygwin AT cygwin DOT com
Message-ID: <cdb1f7f3-e129-08b2-ccd1-467d12de7c9c@t-online.de>
Date: Tue, 25 Mar 2025 14:38:35 +0100
User-Agent: Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:128.0) Gecko/20100101
 SeaMonkey/2.53.20
MIME-Version: 1.0
From: Christian Franke via Cygwin <cygwin AT cygwin DOT com>
Reply-To: cygwin AT cygwin DOT com
Cc: Christian Franke <Christian DOT Franke AT t-online DOT de>
Content-Type: text/plain; charset="utf-8"; Format="flowed"
Sender: "Cygwin" <cygwin-bounces~archive-cygwin=delorie DOT com AT cygwin DOT com>
Content-Transfer-Encoding: 8bit

Found because 'stress-ng --priv-instr ...' hangs and then requires 
'/bin/kill --force ...':

Testcase with
   [PATCH v2] Cygwin: signal: Copy context to alternate stack in the 
SA_ONSTACK case
already applied:

$ uname -r
3.7.0-dev-16-g2ef1a37e7823-dirty.x86_64

$ cat loopsigill.c
#include <setjmp.h>
#include <signal.h>
#include <stdio.h>
#include <unistd.h>

static volatile sig_atomic_t sigcnt;
static sigjmp_buf sjb;

static void sighandler(int sig)
{
   (void)sig;
   ++sigcnt;
   siglongjmp(sjb, 1);
   write(1, "[FAIL]\n", 7);
}

int main()
{
   signal(SIGILL, sighandler);
   printf("pid=%d\n", (int)getpid());

   while (sigsetjmp(sjb, 1))
     ;

   // loop:
   if (sigcnt < 10 || !(sigcnt % 1000))
     printf("%06d\n", sigcnt);
   if (sigcnt >= 100000)
       return 42;
   asm volatile ("invd"); // goto loop;

   return 13; // NOT REACHED
}

$ gcc -o loopsigill loopsigill.c

$ ./loopsigill # may succeed ...
pid=122
000000
000001
...
099000
100000

$ echo $?
42

$ ./loopsigill # ... or crash silently ...
pid=130
000000
000001
...
026000
027000

$ echo $?
0

$ ./loopsigill # ... or hang
pid=135
000000
000001
...
037000
038000
[requires '/bin/kill --force ...']

$ strace -o trace.log ./loopsigill # run '/bin/kill --force ...' ASAP!
pid=142
000000
[always hangs after first signal and fills trace.log quickly]

$ less trace.log
...
    25   25501 [main] loopsigill 142 write: 7 = write(1, 0xA00017710, 7)
--- Process 6856 (pid: 142), exception c0000096 at 00000001004011b9
   142   25643 [main] loopsigill 142 exception::handle: In 
cygwin_except_handler exception 0xC0000096 at 0x1004011B9 sp 0x7FFFFCBE0
    26   25669 [main] loopsigill 142 exception::handle: In 
cygwin_except_handler signal 4 at 0x1004011B9
    38   25707 [main] loopsigill 142 break_here: break here
--- Process 6856 (pid: 142), exception c0000096 at 00000001004011b9
--- Process 6856 (pid: 142), exception c0000096 at 00000001004011b9
... likely repeated until disk is full or time_t wraps around...
--- Process 6856 (pid: 142), exception c0000096 at 00000001004011b9


Problem also occurs
- without the mentioned patch,
- with get/setcontext() instead of sig*jmp(),
- with nullptr access and SIGSEGV handler,
- with Cygwin 3.5.7-1.

I agree that this is not a common use case :-)

-- 
Regards,
Christian


-- 
Problem reports:      https://cygwin.com/problems.html
FAQ:                  https://cygwin.com/faq/
Documentation:        https://cygwin.com/docs.html
Unsubscribe info:     https://cygwin.com/ml/#unsubscribe-simple