Hello Hugo,
"Hugo Heutinck" <hheuti...@mail.altenpts.nl> writes:
> Dear Sir/Madam,
>
> I believe I have detected and a bug in the inetutils tftpd daemon,
> causing a tftpd daemon to hang forever if no ACK is received.
thanks for your report.
It makes sense to replace setjmp/longjmp with the relative signal safe
versions.
Something against this patch?
Cheers,
Giuseppe
>From 20787168168e0a7bec3e51e0f10192ee633a5078 Mon Sep 17 00:00:00 2001
From: Giuseppe Scrivano <gscriv...@gnu.org>
Date: Tue, 11 May 2010 22:01:25 +0200
Subject: [PATCH] Fix a possible hang in tftpd.
---
ChangeLog | 8 ++++++++
src/tftpd.c | 6 +++---
2 files changed, 11 insertions(+), 3 deletions(-)
diff --git a/ChangeLog b/ChangeLog
index 36da17b..0289bf6 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,5 +1,13 @@
2010-05-11 Giuseppe Scrivano <gscriv...@gnu.org>
+ * src/tftpd.c (timer): Use `siglongjmp' instead of `longjmp'.
+ (send_file): Use `sigsetjmp' instead of `setjmp'.
+ (recvfile): Likewise.
+
+ Reported by Hugo Heutinck <hheuti...@mail.altenpts.nl>.
+
+2010-05-11 Giuseppe Scrivano <gscriv...@gnu.org>
+
* src/rcp.c (main): Change the test to check if the call to
`asprintf' failed.
diff --git a/src/tftpd.c b/src/tftpd.c
index 07a8e34..f343f8a 100644
--- a/src/tftpd.c
+++ b/src/tftpd.c
@@ -508,7 +508,7 @@ timer (int sig)
timeout += rexmtval;
if (timeout >= maxtimeout)
exit (1);
- longjmp (timeoutbuf, 1);
+ siglongjmp (timeoutbuf, 1);
}
/*
@@ -537,7 +537,7 @@ send_file (struct formats *pf)
dp->th_opcode = htons ((u_short) DATA);
dp->th_block = htons ((u_short) block);
timeout = 0;
- setjmp (timeoutbuf);
+ sigsetjmp (timeoutbuf, SIGALRM);
send_data:
if (send (peer, (const char *) dp, size + 4, 0) != size + 4)
@@ -608,7 +608,7 @@ recvfile (struct formats *pf)
ap->th_opcode = htons ((u_short) ACK);
ap->th_block = htons ((u_short) block);
block++;
- setjmp (timeoutbuf);
+ sigsetjmp (timeoutbuf, SIGALRM);
send_ack:
if (send (peer, ackbuf, 4, 0) != 4)
{
--
1.7.0
