lzip-bug
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Tarlz 0.26 'make check' test failure on Debian 12


From: Antonio Diaz Diaz
Subject: Re: Tarlz 0.26 'make check' test failure on Debian 12
Date: Wed, 22 Jan 2025 18:20:41 +0100
User-agent: Mozilla/5.0 (X11; U; Linux i586; en-US; rv:1.9.1.19) Gecko/20110420 SeaMonkey/2.0.14

Hello Aren,

Aren Tyr wrote:
I think this is probably just related to the a similar test failure report for 
version 0.25 that was reported on this mailing list, but building tarlz 0.26 
from source and running 'make check' yields the following test failures:

Thank you very much for reporting this.

It looks as if
  touch -d "$i" foo
was not setting 'foo' to the same date as
  tarlz -c --mtime="$i" foo
for dates well outside the 32-bit range.

Apart from that, it seems that tarlz is able to archive all the dates supported by touch on that platform, because the test at line 788 passes.

   0000-01-01T00:00:00foo: Mod time differs

It took me a little more time to find the problem because the /bin/sh (dash?) in Debian 12 does not seem to support the variable LINENO, and does not print the line number of the failed test. It should have printed something like this:

779(0000-01-01T00:00:00)foo: Mod time differs

Meanwhile, I'd just installed/built lzlib-1.15 and lzip-1.25 with no issue 
prior to downloading/building tarlz.

Thanks for the feedback.

I assume this is just a problem with the 'touch' command used in the testsuite 
and not an actual underlying issue with tarlz, i.e. I can safely ignore this 
test failure and use the tool in earnest?

It seems some incompatibility between touch and --mtime for fringe dates. All the other tests pass, so it should be safe to use the compiled tarlz.

I think I'm going to limit the --mtime test to dates since the epoch up to the year 2242 to avoid these incompatibilities with touch.

Best regards,
Antonio.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]