First time building/testing Tcl on a new work system.  It is some version of Ubuntu.  
This is probably a platform bug and not a Tcl bug, but would prefer the test suite not to complain.

==== cmdAH-24.20.1 Tcl_FileObjCmd: atime 64-bit time_t, bug [4718b41c56] FAILED
==== Contents of test case:

    list [file atime $filename 3155760000] [file atime $filename]

---- Result was:
2147483647 2147483647
---- Result should have been (exact matching):
3155760000 3155760000
==== cmdAH-24.20.1 FAILED



==== cmdAH-24.20.2 Tcl_FileObjCmd: mtime 64-bit time_t, bug [4718b41c56] FAILED
==== Contents of test case:

    list [file mtime $filename 3155760000] [file mtime $filename]

---- Result was:
2147483647 2147483647
---- Result should have been (exact matching):
3155760000 3155760000
==== cmdAH-24.20.2 FAILED

Branch tips all show the tests no longer failing.

NP. You could try it again (8.6 only at the moment), [da79c3c0b87a14cc] must fix it (I hope).

Thanks for the followups.  I will engage our local admins to see what they might know.

So then it's probably the disk drive or driver or whatever IO-related there, seeming to have year 2038 problem.

I don't really know how to organize the test suite not to complain here, excepting a constraint using linux commands (touch/stat, but the last with --format is not really posix, imho).

I'll try to provide something here.

% puts pointerSize:\t$::tcl_platform(pointerSize)
pointerSize:	8
% puts st_mtime:\t[testsize st_mtime]
st_mtime:	8

Tests still failing on branch tips.

$ uname -a
Linux HOSTNAME 5.15.0-112-generic #122-Ubuntu SMP Thu May 23 07:48:21 UTC 2024 x86_64 x86_64 x86_64 GNU/Linux

Found a box where it was reproducible previously, and looks OK now, so the constraint work as expected, however only with tcltest-module... it may fail without that, but since it is mostly working (especially on all completely 64-bit platforms), I'll leave it as it is (instead of disabling if no testsize command)...
So I'll close it for now.

Hmm... it looks like continuation of [10aa8403d064439f]... I tried to target there similar issue by extend the constraint for all platforms (previously testsize was for windows only). Seems to be not enough (or we'd need to rewrite the constraint time64bit for some constellation).

The question is what do you see with following lines (in tcltest or in tclsh with loading test-module):

puts pointerSize:\t$::tcl_platform(pointerSize)
puts st_mtime:\t[testsize st_mtime]

I committed my last attempt of more strict attempt in [fd91b0ca09cb171f], would it help?

By the way: is it 32 or 64 bit platform (and kernel)?

Further examination reveals that this is an "Ubuntu 22.04.4 LTS" system.