From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: by sourceware.org (Postfix, from userid 2130) id 49069385415A; Wed, 6 Jul 2022 02:35:24 +0000 (GMT) DKIM-Filter: OpenDKIM Filter v2.11.0 sourceware.org 49069385415A Content-Type: text/plain; charset="us-ascii" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit From: DJ Delorie To: glibc-cvs@sourceware.org Subject: [glibc] test-container: return UNSUPPORTED for ENOSPC on clone() X-Act-Checkin: glibc X-Git-Author: Xi Ruoyao X-Git-Refname: refs/heads/master X-Git-Oldrev: ae308947ff0fa950431d53bcc90c875295d57715 X-Git-Newrev: bd0b58837c7df091046e7531642f379a52e1e157 Message-Id: <20220706023524.49069385415A@sourceware.org> Date: Wed, 6 Jul 2022 02:35:24 +0000 (GMT) X-BeenThere: glibc-cvs@sourceware.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: Glibc-cvs mailing list List-Unsubscribe: , List-Archive: List-Help: List-Subscribe: , X-List-Received-Date: Wed, 06 Jul 2022 02:35:24 -0000 https://sourceware.org/git/gitweb.cgi?p=glibc.git;h=bd0b58837c7df091046e7531642f379a52e1e157 commit bd0b58837c7df091046e7531642f379a52e1e157 Author: Xi Ruoyao Date: Tue Jun 28 18:44:03 2022 +0800 test-container: return UNSUPPORTED for ENOSPC on clone() Since Linux 4.9, the kernel provides /proc/sys/user/max_{mnt,pid,user}_namespace as a limitation of number of namespaces. Some distros (for example, Slint Linux 14.2.1) set them (or only max_user_namespace) to zero as a "security policy" for disabling namespaces. The clone() call will set errno to ENOSPC under such a limitation. We didn't check ENOSPC in the code so the test will FAIL, and report: unable to unshare user/fs: No space left on device This message is, unfortunately, very unhelpful. It leads people to check the memory or disk space, instead of finding the real issue. To improve the situation, we should check for ENOSPC and return UNSUPPORTED as the test result. Also refactor check_for_unshare_hints() to emit a proper message telling people how to make the test work, if they really need to run the namespaced tests. Reported-by: Philippe Delavalade URL: https://lists.linuxfromscratch.org/sympa/arc/lfs-support/2022-06/msg00022.html Signed-off-by: Xi Ruoyao Reviewed-by: DJ Delorie Diff: --- support/test-container.c | 67 ++++++++++++++++++++++++++---------------------- 1 file changed, 36 insertions(+), 31 deletions(-) diff --git a/support/test-container.c b/support/test-container.c index 7557aac441..b6a1158ae1 100644 --- a/support/test-container.c +++ b/support/test-container.c @@ -18,6 +18,7 @@ #define _FILE_OFFSET_BITS 64 +#include #include #include #include @@ -684,39 +685,43 @@ rsync (char *src, char *dest, int and_delete, int force_copies) /* See if we can detect what the user needs to do to get unshare support working for us. */ void -check_for_unshare_hints (void) +check_for_unshare_hints (int require_pidns) { + static struct { + const char *path; + int bad_value, good_value, for_pidns; + } files[] = { + /* Default Debian Linux disables user namespaces, but allows a way + to enable them. */ + { "/proc/sys/kernel/unprivileged_userns_clone", 0, 1, 0 }, + /* ALT Linux has an alternate way of doing the same. */ + { "/proc/sys/kernel/userns_restrict", 1, 0, 0 }, + /* Linux kernel >= 4.9 has a configurable limit on the number of + each namespace. Some distros set the limit to zero to disable the + corresponding namespace as a "security policy". */ + { "/proc/sys/user/max_user_namespaces", 0, 1024, 0 }, + { "/proc/sys/user/max_mnt_namespaces", 0, 1024, 0 }, + { "/proc/sys/user/max_pid_namespaces", 0, 1024, 1 }, + }; FILE *f; - int i; + int i, val; - /* Default Debian Linux disables user namespaces, but allows a way - to enable them. */ - f = fopen ("/proc/sys/kernel/unprivileged_userns_clone", "r"); - if (f != NULL) + for (i = 0; i < array_length (files); i++) { - i = 99; /* Sentinel. */ - fscanf (f, "%d", &i); - if (i == 0) - { - printf ("To enable test-container, please run this as root:\n"); - printf (" echo 1 > /proc/sys/kernel/unprivileged_userns_clone\n"); - } - fclose (f); - return; - } + if (!require_pidns && files[i].for_pidns) + continue; - /* ALT Linux has an alternate way of doing the same. */ - f = fopen ("/proc/sys/kernel/userns_restrict", "r"); - if (f != NULL) - { - i = 99; /* Sentinel. */ - fscanf (f, "%d", &i); - if (i == 1) - { - printf ("To enable test-container, please run this as root:\n"); - printf (" echo 0 > /proc/sys/kernel/userns_restrict\n"); - } - fclose (f); + f = fopen (files[i].path, "r"); + if (f == NULL) + continue; + + val = -1; /* Sentinel. */ + fscanf (f, "%d", &val); + if (val != files[i].bad_value) + continue; + + printf ("To enable test-container, please run this as root:\n"); + printf (" echo %d > %s\n", files[i].good_value, files[i].path); return; } } @@ -1117,11 +1122,11 @@ main (int argc, char **argv) { /* Older kernels may not support all the options, or security policy may block this call. */ - if (errno == EINVAL || errno == EPERM) + if (errno == EINVAL || errno == EPERM || errno == ENOSPC) { int saved_errno = errno; - if (errno == EPERM) - check_for_unshare_hints (); + if (errno == EPERM || errno == ENOSPC) + check_for_unshare_hints (require_pidns); FAIL_UNSUPPORTED ("unable to unshare user/fs: %s", strerror (saved_errno)); } /* We're about to exit anyway, it's "safe" to call unshare again