shaba/systemd-stable
1
1
Fork 0
mirror of https://github.com/systemd/systemd-stable.git synced 2025-01-21 18:03:41 +03:00
Code

Limit rlim_max in rlimit_nofile_safe() to nr_open

Browse Source 
We might inherit a max rlim value that's larger than the kernel's
maximum (nr_open). This will cause setrlimit() to fail as the given
maximum is larger than the kernel's maximum. To get around this,
let's limit the max rlim we pass to rlimit() to the value of nr_open.

Should fix #28965

(cherry picked from commit f470dafddcd688c3ea6031d4bbcbf934fd094711)
(cherry picked from commit a980b83fe07521e4654d68331c31db3a4459412c)
This commit is contained in:
Daan De Meyer 2023-08-25 13:55:36 +02:00 committed by Luca Boccassi
parent b94677ee3d
commit 30ce78cb1a
1 changed files with 5 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

6
src/basic/rlimit-util.c
View File

@ -401,7 +401,11 @@ int rlimit_nofile_safe(void) {
if (rl.rlim_cur <= FD_SETSIZE)
return 0;
rl.rlim_cur = FD_SETSIZE;
/* So we might have inherited a hard limit that's larger than the kernel's maximum limit as stored in
* /proc/sys/fs/nr_open. If we pass this hard limit unmodified to setrlimit(), we'll get EPERM. To
* make sure that doesn't happen, let's limit our hard limit to the value from nr_open. */
rl.rlim_max = MIN(rl.rlim_max, (rlim_t) read_nr_open());
rl.rlim_cur = MIN((rlim_t) FD_SETSIZE, rl.rlim_max);
if (setrlimit(RLIMIT_NOFILE, &rl) < 0)
return log_debug_errno(errno, "Failed to lower RLIMIT_NOFILE's soft limit to " RLIM_FMT ": %m", rl.rlim_cur);