Package: ruby-sigar
Version: 0.7.2-1
Severity: important
Tags: upstream patch
Dear Maintainer,
There is a problem with sigar_proc_port_get function. While looking for
the process that owns a socket, it assumes that the UID of the process
is the same as the UID of the port, which is not always true. I use Chef
to configure some servers, Chef calls Ohai, which uses the function to
generate the port-process map. When a process was started as root,
opened a socket and then dropped priviliges, the function returns 0,
which causes some problems with Ohai and Chef. I tried deleting the
lines related to skipping processes with different UID - after
installing the rebuilt package everything is working fine. I attach
the patch.
Best regards,
Piotr Panczyk
-- System Information:
Debian Release: 7.8
APT prefers oldstable-updates
APT policy: (500, 'oldstable-updates'), (500, 'oldstable')
Architecture: amd64 (x86_64)
Kernel: Linux 2.6.32.bsd54h0 (SMP w/24 CPU cores)
Locale: LANG=en_US.UTF-8, LC_CTYPE=en_US.UTF-8 (charmap=UTF-8)
Shell: /bin/sh linked to /bin/dash
Versions of packages ruby-sigar depends on:
ii libc6 2.13-38+deb7u8
ii libruby1.9.1 1.9.3.194-8.1+deb7u5
ii ruby 1:1.9.3
ii ruby1.9.1 [ruby-interpreter] 1.9.3.194-8.1+deb7u5
ruby-sigar recommends no packages.
ruby-sigar suggests no packages.
-- no debconf information
--- ruby-sigar-0.7.2.orig/src/os/linux/linux_sigar.c
+++ ruby-sigar-0.7.2/src/os/linux/linux_sigar.c
@@ -2502,9 +2502,6 @@ int sigar_proc_port_get(sigar_t *sigar,
if (stat(pid_name, &sb) < 0) {
continue;
}
- if (sb.st_uid != netconn.uid) {
- continue;
- }
/* sprintf(fd_name, "%s/fd", pid_name) */
memcpy(&fd_name[0], pid_name, len);
