wt-status: use capture_command

When we spawn "git submodule status" to read its output, we
use run_command() followed by strbuf_read() read from the
pipe. This can deadlock if the subprocess output is larger
than the system pipe buffer.

Furthermore, if start_command() fails, we'll try to read
from a bogus descriptor (probably "-1" or a descriptor we
just closed, but it is a bad idea for us to make assumptions
about how start_command implements its error handling). And
if start_command succeeds, we leak the file descriptor for
the pipe to the child.

All of these can be solved by using the capture_command
helper.

Signed-off-by: Jeff King <peff@peff.net>
Signed-off-by: Junio C Hamano <gitster@pobox.com>

wt-status.c

@@ -744,11 +744,8 @@ static void wt_status_print_submodule_summary(struct wt_status *s, int uncommitt
 
 	sm_summary.git_cmd = 1;
 	sm_summary.no_stdin = 1;
-	sm_summary.out = -1;
 
-	run_command(&sm_summary);
-
-	strbuf_read(&cmd_stdout, sm_summary.out, 1024);
+	capture_command(&sm_summary, &cmd_stdout, 1024);
 
 	/* prepend header, only if there's an actual output */
 	if (cmd_stdout.len) {