Timestamp out of range; substituting 2514-05-30 01:53:03.999999999

  • Done
  • quality assurance status badge
Details
4 participants
  • Ludovic Courtès
  • Mathieu Othacehe
  • Mark H Weaver
  • Roel Janssen
Owner
unassigned
Submitted by
Roel Janssen
Severity
important
Merged with
R
R
Roel Janssen wrote on 5 Jan 2020 21:45
(address . bug-guix@gnu.org)
d2d7f1ef9036caa9134d29b0227c895511866ec3.camel@gnu.org
Dear Guix,

When inside a container produced with:
$ guix environment -CN --ad-hoc ...

I encounter a problem when running "make" for some code base:
make[3]: /gnu/store/5hkc9q38w6afhrf7xyz5ybxidr87d1mq-profile/include/syslog.h:
Timestamp out of range; substituting 2514-05-30 01:53:03.999999999

This causes subsequent invocations of "make" to (re)compile more than needed.

What can I do about it?

Kind regards,
Roel Janssen
L
L
Ludovic Courtès wrote on 8 Jan 2020 22:58
(name . Roel Janssen)(address . roel@gnu.org)(address . 38958@debbugs.gnu.org)
87o8vdd2zq.fsf@gnu.org
Hi Roel,

Roel Janssen <roel@gnu.org> skribis:

Toggle quote (7 lines)
> When inside a container produced with:
> $ guix environment -CN --ad-hoc ...
>
> I encounter a problem when running "make" for some code base:
> make[3]: /gnu/store/5hkc9q38w6afhrf7xyz5ybxidr87d1mq-profile/include/syslog.h:
> Timestamp out of range; substituting 2514-05-30 01:53:03.999999999

Fun. :-) What does “uname -rm” return?

What the value of the ‘TZ’ environment variable?

Can it be easily reproduced?

Thanks,
Ludo’.
L
L
Ludovic Courtès wrote on 27 Jan 2020 22:50
control message for bug #38958
(address . control@debbugs.gnu.org)
87blqofu0t.fsf@gnu.org
tags 38958 + moreinfo
quit
R
R
Roel Janssen wrote on 27 Mar 2020 10:26
Re: bug#38958: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
(name . Ludovic Courtès)(address . ludo@gnu.org)(address . 38958@debbugs.gnu.org)
8aab63ee4ca678aab40be28edbc6f3178d1c8acb.camel@gnu.org
On Wed, 2020-01-08 at 22:58 +0100, Ludovic Courtès wrote:
Toggle quote (22 lines)
> Hi Roel,
>
> Roel Janssen <roel@gnu.org> skribis:
>
> > When inside a container produced with:
> > $ guix environment -CN --ad-hoc ...
> >
> > I encounter a problem when running "make" for some code base:
> > make[3]: /gnu/store/5hkc9q38w6afhrf7xyz5ybxidr87d1mq-
> > profile/include/syslog.h:
> > Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
>
> Fun. :-) What does “uname -rm” return?
>
> What the value of the ‘TZ’ environment variable?
>
> Can it be easily reproduced?
>
> Thanks,
> Ludo’.
>

Sorry for the long delay. This issue had somehow resolved itself.

But now the issue resurfaced for me when building an older version of QtSvg
(5.9.4). The 'TZ' environment variable is empty, and the output of "uname -rm"
is:
5.5.10-200.fc31.x86_64 x86_64

The package recipe is somewhat involved (it's part of a build for RStudio), but
if you'd like I can attach it here.

Kind regards,
Roel Janssen
L
L
Ludovic Courtès wrote on 27 Mar 2020 12:54
(name . Roel Janssen)(address . roel@gnu.org)(address . 38958@debbugs.gnu.org)
87v9mq57z5.fsf@gnu.org
Roel Janssen <roel@gnu.org> skribis:

Toggle quote (30 lines)
> On Wed, 2020-01-08 at 22:58 +0100, Ludovic Courtès wrote:
>> Hi Roel,
>>
>> Roel Janssen <roel@gnu.org> skribis:
>>
>> > When inside a container produced with:
>> > $ guix environment -CN --ad-hoc ...
>> >
>> > I encounter a problem when running "make" for some code base:
>> > make[3]: /gnu/store/5hkc9q38w6afhrf7xyz5ybxidr87d1mq-
>> > profile/include/syslog.h:
>> > Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
>>
>> Fun. :-) What does “uname -rm” return?
>>
>> What the value of the ‘TZ’ environment variable?
>>
>> Can it be easily reproduced?
>>
>> Thanks,
>> Ludo’.
>>
>
> Sorry for the long delay. This issue had somehow resolved itself.
>
> But now the issue resurfaced for me when building an older version of QtSvg
> (5.9.4). The 'TZ' environment variable is empty, and the output of "uname -rm"
> is:
> 5.5.10-200.fc31.x86_64 x86_64

Previously you mentioned the problem was in ‘guix environment -C’, but
now you say it happens while running ‘guix build qtsvg’, right?

Is there a simple way to reproduce it?

Ludo’.
L
L
Ludovic Courtès wrote on 3 Sep 2020 10:54
(address . 38958@debbugs.gnu.org)
87lfhr1bra.fsf@gnu.org
Hi,

Roel Janssen <roel@gnu.org> skribis:

Toggle quote (4 lines)
> I encounter a problem when running "make" for some code base:
> make[3]: /gnu/store/5hkc9q38w6afhrf7xyz5ybxidr87d1mq-profile/include/syslog.h:
> Timestamp out of range; substituting 2514-05-30 01:53:03.999999999

This issue manifests itself in build environments as well; I’ve seen it
on berlin while running “guix build guix --no-offload”:

Toggle snippet (40 lines)
mv "doc/contributing.fr.texi.tmp" "doc/contributing.fr.texi"
PO4A doc/contributing.zh_CN.texi
mv "doc/contributing.ru.texi.tmp" "doc/contributing.ru.texi"
make: guix/store/schema.sql: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make: Warning: File 'guix/store/schema.sql' has modification time 15580747586 s in the future
PO4A doc/guix-cookbook.de.texi
[…]
test ! -f guix.po || { \
if test -f ./guix.pot; then \
sed -f remove-potcdate.sed < ./guix.pot > guix.1po && \
sed -f remove-potcdate.sed < guix.po > guix.2po && \
if cmp guix.1po guix.2po >/dev/null 2>&1; then \
rm -f guix.1po guix.2po guix.po; \
else \
rm -f guix.1po guix.2po ./guix.pot && \
mv guix.po ./guix.pot; \
fi; \
else \
mv guix.po ./guix.pot; \
fi; \
}
make[3]: warning: Clock skew detected. Your build may be incomplete.
make[3]: Leaving directory '/tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/po/guix'
[…]
make[2]: guix/base16.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: Warning: File 'guix/base16.scm' has modification time 15580748038 s in the future
make[2]: guix/base32.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
[…]
make[2]: guix/search-paths.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/import/gnu.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/import/snix.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/scripts/perform-download.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/scripts/authenticate.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/scripts/import/gnu.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/scripts/import/nix.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/scripts/import/texlive.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix/scripts/container.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
make[2]: guix.scm: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999

The files in question have mtime = Epoch + 1 as expected:

Toggle snippet (37 lines)
$ stat /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/base16.scm
File: /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/base16.scm
Size: 3251 Blocks: 8 IO Block: 4096 regular file
Device: 801h/2049d Inode: 10880988 Links: 1
Access: (0444/-r--r--r--) Uid: (30001/guixbuilder01) Gid: (30000/guixbuild)
Access: 2020-09-03 10:18:35.594069572 +0200
Modify: 1970-01-01 00:00:01.000000000 +0100
Change: 2020-09-03 10:18:27.670122206 +0200
Birth: 2020-09-03 10:18:27.670122206 +0200
$ stat /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/base32.scm
File: /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/base32.scm
Size: 13325 Blocks: 32 IO Block: 4096 regular file
Device: 801h/2049d Inode: 10881034 Links: 1
Access: (0444/-r--r--r--) Uid: (30001/guixbuilder01) Gid: (30000/guixbuild)
Access: 2020-09-03 10:18:35.594069572 +0200
Modify: 1970-01-01 00:00:01.000000000 +0100
Change: 2020-09-03 10:18:27.674122179 +0200
Birth: 2020-09-03 10:18:27.674122179 +0200
$ stat /gnu/store/0br13w7y2wdfby74wnq7m7yzbx27hm1s-guix-1.1.0-25.44c6e6f-checkout/guix/base{16,32}.scm
File: /gnu/store/0br13w7y2wdfby74wnq7m7yzbx27hm1s-guix-1.1.0-25.44c6e6f-checkout/guix/base16.scm
Size: 3251 Blocks: 8 IO Block: 4096 regular file
Device: 811h/2065d Inode: 344654478 Links: 10338
Access: (0444/-r--r--r--) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2020-09-03 09:36:43.818923620 +0200
Modify: 1970-01-01 00:00:01.000000000 +0100
Change: 2020-09-03 09:29:05.830024528 +0200
Birth: 2018-06-25 22:37:25.560482991 +0200
File: /gnu/store/0br13w7y2wdfby74wnq7m7yzbx27hm1s-guix-1.1.0-25.44c6e6f-checkout/guix/base32.scm
Size: 13325 Blocks: 32 IO Block: 4096 regular file
Device: 811h/2065d Inode: 344654504 Links: 9327
Access: (0444/-r--r--r--) Uid: ( 0/ root) Gid: ( 0/ root)
Access: 2020-09-03 09:36:43.822923593 +0200
Modify: 1970-01-01 00:00:01.000000000 +0100
Change: 2020-09-03 09:29:07.054016225 +0200
Birth: 2018-06-25 22:37:25.682483782 +0200

In the output of ‘make’ above, guix/import/gnu.scm is listed, but not
guix/import/pypi.scm for instance. What’s the difference?

Toggle snippet (19 lines)
$ stat /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/import/{gnu,pypi}.scm
File: /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/import/gnu.scm
Size: 4714 Blocks: 16 IO Block: 4096 regular file
Device: 801h/2049d Inode: 10881084 Links: 1
Access: (0444/-r--r--r--) Uid: (30001/guixbuilder01) Gid: (30000/guixbuild)
Access: 2020-09-03 10:26:30.745348482 +0200
Modify: 1970-01-01 00:00:01.000000000 +0100
Change: 2020-09-03 10:26:23.029400312 +0200
Birth: 2020-09-03 10:26:23.029400312 +0200
File: /tmp/guix-build-guix-1.1.0-25.44c6e6f.drv-0/source/guix/import/pypi.scm
Size: 23270 Blocks: 48 IO Block: 4096 regular file
Device: 801h/2049d Inode: 10881093 Links: 1
Access: (0444/-r--r--r--) Uid: (30001/guixbuilder01) Gid: (30000/guixbuild)
Access: 2020-09-03 10:26:30.745348482 +0200
Modify: 1970-01-01 01:00:01.000000000 +0100
Change: 2020-09-03 10:26:23.029400312 +0200
Birth: 2020-09-03 10:26:23.029400312 +0200

Nothing obvious.

After checking the clock was fine, I turned off ntpd to be sure and
spawned another build, but the result was the same.

This is happening on berlin with a recent kernel:

Toggle snippet (4 lines)
$ uname -rm
5.4.55-gnu x86_64

The GNU make warnings come from this impenetrable function:

Toggle snippet (23 lines)
FILE_TIMESTAMP
file_timestamp_cons (const char *fname, time_t stamp, long int ns)
{
int offset = ORDINARY_MTIME_MIN + (FILE_TIMESTAMP_HI_RES ? ns : 0);
FILE_TIMESTAMP s = stamp;
FILE_TIMESTAMP product = (FILE_TIMESTAMP) s << FILE_TIMESTAMP_LO_BITS;
FILE_TIMESTAMP ts = product + offset;

if (! (s <= FILE_TIMESTAMP_S (ORDINARY_MTIME_MAX)
&& product <= ts && ts <= ORDINARY_MTIME_MAX))
{
char buf[FILE_TIMESTAMP_PRINT_LEN_BOUND + 1];
const char *f = fname ? fname : _("Current time");
ts = s <= OLD_MTIME ? ORDINARY_MTIME_MIN : ORDINARY_MTIME_MAX;
file_timestamp_sprintf (buf, ts);
OSS (error, NILF,
_("%s: Timestamp out of range; substituting %s"), f, buf);
}

return ts;
}

What’s OLD_MTIME?

Toggle snippet (8 lines)
/* The file does not exist, and we assume that it is older than any
actual file. */
#define OLD_MTIME 2

/* The smallest and largest ordinary timestamps. */
#define ORDINARY_MTIME_MIN (OLD_MTIME + 1)

That would mean that any file with mtime < 3 is considered bogus, but
then, why wouldn’t things fail on other machines as well?

I’m looking for ideas! :-)

Ludo’.
M
M
Mathieu Othacehe wrote on 3 Sep 2020 13:54
control message for bug #38958
(address . control@debbugs.gnu.org)
878sdrdqjc.fsf@cervin.i-did-not-set--mail-host-address--so-tickle-me
merge 38958 42148
quit
M
M
Mark H Weaver wrote on 3 Sep 2020 21:42
Re: bug#38958: Timestamp out of range; substituting 2514-05-30 01:53:03.999999999
87y2lq3avu.fsf@netris.org
Hi,

Ludovic Courtès <ludo@gnu.org> writes:

Toggle quote (40 lines)
> The GNU make warnings come from this impenetrable function:
>
> --8<---------------cut here---------------start------------->8---
> FILE_TIMESTAMP
> file_timestamp_cons (const char *fname, time_t stamp, long int ns)
> {
> int offset = ORDINARY_MTIME_MIN + (FILE_TIMESTAMP_HI_RES ? ns : 0);
> FILE_TIMESTAMP s = stamp;
> FILE_TIMESTAMP product = (FILE_TIMESTAMP) s << FILE_TIMESTAMP_LO_BITS;
> FILE_TIMESTAMP ts = product + offset;
>
> if (! (s <= FILE_TIMESTAMP_S (ORDINARY_MTIME_MAX)
> && product <= ts && ts <= ORDINARY_MTIME_MAX))
> {
> char buf[FILE_TIMESTAMP_PRINT_LEN_BOUND + 1];
> const char *f = fname ? fname : _("Current time");
> ts = s <= OLD_MTIME ? ORDINARY_MTIME_MIN : ORDINARY_MTIME_MAX;
> file_timestamp_sprintf (buf, ts);
> OSS (error, NILF,
> _("%s: Timestamp out of range; substituting %s"), f, buf);
> }
>
> return ts;
> }
> --8<---------------cut here---------------end--------------->8---
>
> What’s OLD_MTIME?
>
> --8<---------------cut here---------------start------------->8---
> /* The file does not exist, and we assume that it is older than any
> actual file. */
> #define OLD_MTIME 2
>
> /* The smallest and largest ordinary timestamps. */
> #define ORDINARY_MTIME_MIN (OLD_MTIME + 1)
> --8<---------------cut here---------------end--------------->8---
>
> That would mean that any file with mtime < 3 is considered bogus, but
> then, why wouldn’t things fail on other machines as well?

I spent a bit of time looking at the relevant code in GNU Make. The
special MTIME values of 0, 1, and 2 seem to apply only to GNU Make's
*internal* representation of the timestamp. 'file_timestamp_cons',
which converts a standard POSIX time to the internal representation,
seems to properly handle times near the POSIX epoch by adding
ORDINARY_MTIME_MIN (via 'offset') to the POSIX time, after multiplying
it by 2^30 (if FILE_TIMESTAMP_HI_RES is enabled).

Toggle quote (2 lines)
> I’m looking for ideas! :-)

Note that the date printed in the warning (ORDINARY_MTIME_MAX),
represented as a POSIX time (seconds past the epoch), is precisely 2^34
seconds minus one nanosecond.

The problem doesn't seem to be that 'stamp' is too small, because if it
were, then the following line in 'file_timestamp_cons',

ts = s <= OLD_MTIME ? ORDINARY_MTIME_MIN : ORDINARY_MTIME_MAX;

would substitute ORDINARY_MTIME_MIN, which is close to the POSIX epoch,
and the warning message would print a time near 1970, instead of one
near 2514 (ORDINARY_MTIME_MAX).

Rather, it appears that the 'stamp' passed into 'file_timestamp_cons'
was close to or larger than 2^34, which is approximately the largest
timestamp that GNU make supports when FILE_TIMESTAMP is 64 bits and
FILE_TIMESTAMP_HI_RES is enabled.

My guess is that maybe our near-zero timestamps are somewhere being
adjusted downwards by a timezone conversion, using an unsigned integer
type, causing them to wrap around to near the maximum value of that
type.

Note that although 'stamp' usually comes from a file 'mtime' as returned
by stat(2), it can also come from an 'ar' archive member. In
make-4.3/src/remake.c, 'f_mtime' includes the following code:

Toggle snippet (6 lines)
member_date = ar_member_date (file->hname);
mtime = (member_date == (time_t) -1
? NONEXISTENT_MTIME
: file_timestamp_cons (file->hname, member_date, 0));

Mark
L
L
Ludovic Courtès wrote on 16 Sep 2020 21:57
control message for bug #38958
(address . control@debbugs.gnu.org)
871rj1mr3v.fsf@gnu.org
severity 38958 important
quit
?
Your comment

This issue is archived.

To comment on this conversation send an email to 38958@debbugs.gnu.org

To respond to this issue using the mumi CLI, first switch to it
mumi current 38958
Then, you may apply the latest patchset in this issue (with sign off)
mumi am -- -s
Or, compose a reply to this issue
mumi compose
Or, send patches to this issue
mumi send-email *.patch