Re: [Qemu-devel] [PATCH] virtfs-proxy-helper: check return code of setfsgid/setfsuid

[Stefan Weil]

Am 10.10.2012 13:32, schrieb Paolo Bonzini:

Fixes the following error with glibc 2.16 on Fedora 18:

virtfs-proxy-helper.c: In function ‘setfsugid’:
virtfs-proxy-helper.c:293:13: error: ignoring return value of ‘setfsgid’, 
declared with attribute warn_unused_result [-Werror=unused-result]
virtfs-proxy-helper.c:294:13: error: ignoring return value of ‘setfsuid’, 
declared with attribute warn_unused_result [-Werror=unused-result]
cc1: all warnings being treated as errors

Signed-off-by: Paolo Bonzini <pbonz...@redhat.com>
---
  fsdev/virtfs-proxy-helper.c | 8 ++++++--
  1 file modificato, 6 inserzioni(+), 2 rimozioni(-)

diff --git a/fsdev/virtfs-proxy-helper.c b/fsdev/virtfs-proxy-helper.c
index f9a8270..b34a84a 100644
--- a/fsdev/virtfs-proxy-helper.c
+++ b/fsdev/virtfs-proxy-helper.c
@@ -290,8 +290,12 @@ static int setfsugid(int uid, int gid)
          CAP_DAC_OVERRIDE,
      };

- setfsgid(gid);

-    setfsuid(uid);
+    if (setfsgid(gid) != 0) {
+        return -1;
+    }

Wouldn't setfsgid(gid) == gid be also ok?

I have no idea whether it is a realistic scenario to call
that function with a gid which is already set.

+    if (setfsuid(uid) != 0) {
+        return -1;
+    }

The same question applies here.

      if (uid != 0 || gid != 0) {
          return do_cap_set(cap_list, ARRAY_SIZE(cap_list), 0);

The Linux manpage for both functions says something
completely different. Extract from manpage of setfsuid:

"On success, the previous value of fsuid is returned.
 On error, the current value of fsuid is returned."

In reality, the functions return 0 when called by root.

Regards

Stefan