Fix endtime issue

Fix endtime issue

[Doug Nazar]

[-- Attachment #1: Type: text/plain, Size: 282 bytes --]

  After finally tracking down an issue with rpc.gssd crashing which had 
been bugging me on and off for months, ran into another issue.

Commit 4c5ff6d4 removed the setting of endtime for a few contexts by 
accident.

Now to figure out why I get stale handles on submounts.

Doug



[-- Attachment #2: nfs-utils-endtime.diff --]
[-- Type: text/plain, Size: 1327 bytes --]

commit 3e4355dd4a31ee8577620b8b93cd5a811cdcaa7c
Author: Doug Nazar <nazard.michi@gmail.com>
Date:   Wed Jun 30 15:49:39 2010 -0400

    Return correct context endtime for lucid rfc4121 and mit.

diff --git a/utils/gssd/context_lucid.c b/utils/gssd/context_lucid.c
index b87bf76..2108487 100644
--- a/utils/gssd/context_lucid.c
+++ b/utils/gssd/context_lucid.c
@@ -202,6 +202,8 @@ prepare_krb5_rfc4121_buffer(gss_krb5_lucid_context_v1_t *lctx,
 
 	if (WRITE_BYTES(&p, end, v2_flags)) goto out_err;
 	if (WRITE_BYTES(&p, end, lctx->endtime)) goto out_err;
+	if (endtime)
+		*endtime = lctx->endtime;
 	if (WRITE_BYTES(&p, end, lctx->send_seq)) goto out_err;
 
 	/* Protocol 0 here implies DES3 or RC4 */
diff --git a/utils/gssd/context_mit.c b/utils/gssd/context_mit.c
index f9cbb02..e6db9cb 100644
--- a/utils/gssd/context_mit.c
+++ b/utils/gssd/context_mit.c
@@ -191,6 +191,8 @@ serialize_krb5_ctx(gss_ctx_id_t ctx, gss_buffer_desc *buf, int32_t *endtime)
 		if (WRITE_BYTES(&p, end, kctx->signalg)) goto out_err;
 		if (WRITE_BYTES(&p, end, kctx->sealalg)) goto out_err;
 		if (WRITE_BYTES(&p, end, kctx->endtime)) goto out_err;
+		if (endtime)
+			*endtime = kctx->endtime;
 		word_seq_send = kctx->seq_send;
 		if (WRITE_BYTES(&p, end, word_seq_send)) goto out_err;
 		if (write_oid(&p, end, kctx->mech_used)) goto out_err;

Re: Fix endtime issue

[Steve Dickson]

On 06/30/2010 09:04 PM, Doug Nazar wrote:
>  After finally tracking down an issue with rpc.gssd crashing which had
> been bugging me on and off for months, ran into another issue.
> 
> Commit 4c5ff6d4 removed the setting of endtime for a few contexts by
> accident.
> 
> Now to figure out why I get stale handles on submounts.
Committed...

steved.