[PULL,1/1] seccomp: fix killing of whole process instead of thread

Commit Message

Eduardo Otubo Sept. 15, 2020, 11:01 a.m. UTC

From: Daniel P. Berrangé <[e-mail]>

Back in 2018 we introduced support for killing the whole QEMU process
instead of just one thread, when a seccomp rule is violated:

  commit bda08a5764d470f101fa38635d30b41179a313e1
  Author: Marc-André Lureau <[e-mail]>
  Date:   Wed Aug 22 19:02:48 2018 +0200

    seccomp: prefer SCMP_ACT_KILL_PROCESS if available

Fast forward a year and we introduced a patch to avoid killing the
process for resource control syscalls tickled by Mesa.

  commit 9a1565a03b79d80b236bc7cc2dbce52a2ef3a1b8
  Author: Daniel P. Berrangé <[e-mail]>
  Date:   Wed Mar 13 09:49:03 2019 +0000

    seccomp: don't kill process for resource control syscalls

Unfortunately a logic bug effectively reverted the first commit
mentioned so that we go back to only killing the thread, not the whole
process.

Signed-off-by: Daniel P. Berrangé <[e-mail]>
Reviewed-by: Stefan Hajnoczi <[e-mail]>
Acked-by: Eduardo Otubo <[e-mail]>
---
 qemu-seccomp.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Patch

diff --git a/qemu-seccomp.c b/qemu-seccomp.c
index e0a1829b3d..8325ecb766 100644
--- a/qemu-seccomp.c
+++ b/qemu-seccomp.c
@@ -136,8 +136,9 @@  static uint32_t qemu_seccomp_get_action(int set)
 
             if (qemu_seccomp(SECCOMP_GET_ACTION_AVAIL, 0, &action) == 0) {
                 kill_process = 1;
+            } else {
+                kill_process = 0;
             }
-            kill_process = 0;
         }
         if (kill_process == 1) {
             return SCMP_ACT_KILL_PROCESS;