From f5c60ff2da4851f9056120a423ce6b48624fb97e Mon Sep 17 00:00:00 2001
From: Alexander Morozov <lk4d4math@gmail.com>
Date: Wed, 3 Jun 2015 10:50:39 -0700
Subject: syscall: add GidMappingsEnableSetgroups to Linux SysProcAttr

Linux 3.19 made a change in the handling of setgroups and the 'gid_map' file to
address a security issue.
The upshot of the 3.19 changes is that in order to update the 'gid_maps' file,
use of the setgroups() system call in this user namespace must first be disabled
by writing "deny" to one of the /proc/PID/setgroups files for this namespace.

Also added tests for remapping uid_map and gid_map inside new user
namespace.

Fixes #10626

Change-Id: I4d2539acbab741a37092d277e10f31fc39a8feb7
Reviewed-on: https://go-review.googlesource.com/10670
Run-TryBot: Ian Lance Taylor <iant@golang.org>
Reviewed-by: Ian Lance Taylor <iant@golang.org>
---
 src/syscall/exec_linux.go | 35 +++++++++++++++++++++++++++++++++++
 1 file changed, 35 insertions(+)

(limited to 'src/syscall/exec_linux.go')

diff --git a/src/syscall/exec_linux.go b/src/syscall/exec_linux.go
index 3aa30c7364..9bac042124 100644
--- a/src/syscall/exec_linux.go
+++ b/src/syscall/exec_linux.go
@@ -33,6 +33,11 @@ type SysProcAttr struct {
 	Cloneflags  uintptr        // Flags for clone calls (Linux only)
 	UidMappings []SysProcIDMap // User ID mappings for user namespaces.
 	GidMappings []SysProcIDMap // Group ID mappings for user namespaces.
+	// GidMappingsEnableSetgroups enabling setgroups syscall.
+	// If false, then setgroups syscall will be disabled for the child process.
+	// This parameter is no-op if GidMappings == nil. Otherwise for unprivileged
+	// users this should be set to false for mappings work.
+	GidMappingsEnableSetgroups bool
 }
 
 // Implemented in runtime package.
@@ -366,6 +371,32 @@ func writeIDMappings(path string, idMap []SysProcIDMap) error {
 	return nil
 }
 
+// writeSetgroups writes to /proc/PID/setgroups "deny" if enable is false
+// and "allow" if enable is true.
+// This is needed since kernel 3.19, because you can't write gid_map without
+// disabling setgroups() system call.
+func writeSetgroups(pid int, enable bool) error {
+	sgf := "/proc/" + itoa(pid) + "/setgroups"
+	fd, err := Open(sgf, O_RDWR, 0)
+	if err != nil {
+		return err
+	}
+
+	var data []byte
+	if enable {
+		data = []byte("allow")
+	} else {
+		data = []byte("deny")
+	}
+
+	if _, err := Write(fd, data); err != nil {
+		Close(fd)
+		return err
+	}
+
+	return Close(fd)
+}
+
 // writeUidGidMappings writes User ID and Group ID mappings for user namespaces
 // for a process and it is called from the parent process.
 func writeUidGidMappings(pid int, sys *SysProcAttr) error {
@@ -377,6 +408,10 @@ func writeUidGidMappings(pid int, sys *SysProcAttr) error {
 	}
 
 	if sys.GidMappings != nil {
+		// If the kernel is too old to support /proc/PID/setgroups, writeSetGroups will return ENOENT; this is OK.
+		if err := writeSetgroups(pid, sys.GidMappingsEnableSetgroups); err != nil && err != ENOENT {
+			return err
+		}
 		gidf := "/proc/" + itoa(pid) + "/gid_map"
 		if err := writeIDMappings(gidf, sys.GidMappings); err != nil {
 			return err
-- 
cgit v1.3