prctl.2: PR_SET_SECCOMP - _exit behavior

[Ivana Hutarova Varekova <varekova-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>]

[-- Attachment #1: Type: text/plain, Size: 507 bytes --]

prctl.2 man page (man-pages-3.30) wrote:
  PR_SET_SECCOMP ... After the secure computing mode has been set to 1, 
the only system calls that the thread is permitted to make are read(2), 
write(2),  _exit(2), and sigreturn(2),
but _exit(2) wrapper from 2.3 does not call _exit() sysscall but invokes 
exit_group(2) syscall(see exit.2). This may lead to misunderstanding. 
prctl.2 man page may contain some note about this fact too.
The patch which adds this note to prctl is attached
Ivana HUtarova Varekova

[-- Attachment #2: man-pages-3.30-prctl.patch --]
[-- Type: text/x-patch, Size: 695 bytes --]

diff -up man-pages-3.30/man2/prctl.2.pom man-pages-3.30/man2/prctl.2
--- man-pages-3.30/man2/prctl.2.pom	2010-11-01 15:55:38.000000000 +0100
+++ man-pages-3.30/man2/prctl.2	2010-11-18 14:42:50.000000000 +0100
@@ -236,9 +236,15 @@ After the secure computing mode has been
 the only system calls that the thread is permitted to make are
 .BR read (2),
 .BR write (2),
-.BR _exit (2),
+.BR sigreturn (2)
 and
-.BR sigreturn (2).
+.BR _exit (2).
+In glibc up to version 2.3, the
+.BR _exit()
+wrapper function invoked the kernel system call of the same name.
+Since glibc 2.3, the  wrapper  function
+invokes
+.BR exit_group(2).
 Other system calls result in the delivery of a
 .BR SIGKILL
 signal.