delorie.com/archives/browse.cgi | search |
X-Recipient: | archive-cygwin AT delorie DOT com |
X-SWARE-Spam-Status: | No, hits=1.1 required=5.0 tests=BAYES_05,FREEMAIL_ENVFROM_END_DIGIT,FREEMAIL_FROM,RCVD_IN_DNSWL_NONE,T_TO_NO_BRKTS_FREEMAIL |
X-Spam-Check-By: | sourceware.org |
Message-Id: | <E53F32EF-0D79-4541-B483-3405336B2FC1@free.fr> |
From: | Denis Excoffier <3 DOT 1416 AT free DOT fr> |
To: | cygwin AT cygwin DOT com |
Reply-To: | cygwin AT denis-excoffier DOT org |
Subject: | unexpected jumps to the future in the system date |
Mime-Version: | 1.0 (Apple Message framework v936) |
Date: | Sun, 19 Dec 2010 22:13:53 +0100 |
Cc: | lasse DOT collin AT tukaani DOT org, cygwin AT denis-excoffier DOT org |
Mailing-List: | contact cygwin-help AT cygwin DOT com; run by ezmlm |
List-Id: | <cygwin.cygwin.com> |
List-Subscribe: | <mailto:cygwin-subscribe AT cygwin DOT com> |
List-Archive: | <http://sourceware.org/ml/cygwin/> |
List-Post: | <mailto:cygwin AT cygwin DOT com> |
List-Help: | <mailto:cygwin-help AT cygwin DOT com>, <http://sourceware.org/ml/#faqs> |
Sender: | cygwin-owner AT cygwin DOT com |
Mail-Followup-To: | cygwin AT cygwin DOT com |
Delivered-To: | mailing list cygwin AT cygwin DOT com |
Hello, I've experienced a curious behaviour of /usr/bin/date, more especially some large jumps in the future. Since i cannot formally exclude a BLODA problem, i'm not very comfortable at sending this, but see the additional remarks at the end. Program is `mytest': ---8<--------------------- #!/bin/sh rm -f f1 f1.xz echo "dummy text" > f1 # next line removes f1 and creates f1.xz (for those unfamiliar with `xz') /usr/bin/xz -8 f1 /usr/sbin/zdump UTC /usr/bin/date -u /usr/bin/xz -d < f1.xz > /dev/null /usr/bin/date -u /usr/sbin/zdump UTC rm -f f1 f1.xz ---8<---------------------- Result is (lines starting by # have been added by me afterwards): --------------------------- Script started on Fri Dec 17 17:49:56 2010 jupiter% mytest # time given by `zdump` is always correct UTC Fri Dec 17 16:50:00 2010 UTC # time given by `date` is still correct here Fri Dec 17 16:50:00 UTC 2010 # # for `date`, there is a jump of 4.5 days! # Wed Dec 22 03:17:46 UTC 2010 # time given by `zdump` is always correct UTC Fri Dec 17 16:50:00 2010 UTC jupiter% mytest # # 5 seconds later... UTC Fri Dec 17 16:50:05 2010 UTC # `date` also is 5 seconds later Wed Dec 22 03:17:51 UTC 2010 # # now a jump of 6 seconds! # Wed Dec 22 03:17:57 UTC 2010 # in fact, less than 1 second has elapsed in reality UTC Fri Dec 17 16:50:05 2010 UTC jupiter% mytest # # again 5 seconds later UTC Fri Dec 17 16:50:10 2010 UTC # `date` also is 5 seconds later Wed Dec 22 03:18:02 UTC 2010 # # jump of 5 seconds! # Wed Dec 22 03:18:07 UTC 2010 # in fact, less than 1 second has elapsed in reality UTC Fri Dec 17 16:50:10 2010 UTC jupiter% exit exit Script done on Fri Dec 17 17:50:11 2010 --------------------------- Remarks: - i use XP, with McAfee perhaps not completely deactivated - all versions are the last available in the mirrors: /usr/bin/xz, /usr/bin/zdump, /usr/bin/date, cygwin1.dll - /usr/sbin/zdump is simply an alternate command to print the date it seems to remain functioning in the storm - the -8 is an important detail: 32Mo will be needed to decompress the file; a similar option (eg -7 or -9) does not produce any weird behaviour - i tried to replace the `xz -d' line with `dd bs=33M count=1 < /dev/zero > /dev/null' or similar but nothing special happened; also `xz -cd f1.xz > /dev/null' does not work - this is 100% reproducible, only the jump amounts seems erratic -- Problem reports: http://cygwin.com/problems.html FAQ: http://cygwin.com/faq/ Documentation: http://cygwin.com/docs.html Unsubscribe info: http://cygwin.com/ml/#unsubscribe-simple
webmaster | delorie software privacy |
Copyright © 2019 by DJ Delorie | Updated Jul 2019 |