GNU bug report logs - #48131
build failure: 1.3.0rc1 on Debian i386 with guile-2.2

Previous Next

Package: guix;

Reported by: Vagrant Cascadian <vagrant <at> debian.org>

Date: Sat, 1 May 2021 00:03:02 UTC

Severity: normal

Done: Vagrant Cascadian <vagrant <at> debian.org>

Bug is archived. No further changes may be made.

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 48131 in the body.
You can then email your comments to 48131 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

View this report as an mbox folder, status mbox, maintainer mbox


Report forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Sat, 01 May 2021 00:03:02 GMT) Full text and rfc822 format available.

Acknowledgement sent to Vagrant Cascadian <vagrant <at> debian.org>:
New bug report received and forwarded. Copy sent to bug-guix <at> gnu.org. (Sat, 01 May 2021 00:03:02 GMT) Full text and rfc822 format available.

Message #5 received at submit <at> debbugs.gnu.org (full text, mbox):

From: Vagrant Cascadian <vagrant <at> debian.org>
To: bug-guix <at> gnu.org
Subject: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Fri, 30 Apr 2021 17:01:59 -0700
[Message part 1 (text/plain, inline)]
I've been unable to build version 1.3.0rc1 for i386 on Debian:

  [ 70%] GUILEC   gnu/packages/direct-connect.go
  [ 70%] GUILEC   gnu/packages/disk.go
  GC Warning: Failed to expand heap by 8388608 bytes
  GC Warning: Failed to expand heap by 8388608 bytes
  ...
  GC Warning: Failed to expand heap by 8388608 bytes
  GC Warning: Failed to expand heap by 8388608 bytes
  Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS
  make[3]: *** [Makefile:6871: make-go] Aborted


Debian is still using guile-2.2, so I'm guessing this is not a
particularly well tested codepath anymore...


live well,
  vagrant
[signature.asc (application/pgp-signature, inline)]

Information forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Sun, 02 May 2021 02:18:02 GMT) Full text and rfc822 format available.

Message #8 received at 48131 <at> debbugs.gnu.org (full text, mbox):

From: Vagrant Cascadian <vagrant <at> debian.org>
To: 48131 <at> debbugs.gnu.org
Subject: Re: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Sat, 01 May 2021 19:17:12 -0700
[Message part 1 (text/plain, inline)]
On 2021-04-30, Vagrant Cascadian wrote:
> I've been unable to build version 1.3.0rc1 for i386 on Debian:
>
>   [ 70%] GUILEC   gnu/packages/direct-connect.go
>   [ 70%] GUILEC   gnu/packages/disk.go
>   GC Warning: Failed to expand heap by 8388608 bytes
>   GC Warning: Failed to expand heap by 8388608 bytes
>   ...
>   GC Warning: Failed to expand heap by 8388608 bytes
>   GC Warning: Failed to expand heap by 8388608 bytes
>   Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS
>   make[3]: *** [Makefile:6871: make-go] Aborted
>
>
> Debian is still using guile-2.2, so I'm guessing this is not a
> particularly well tested codepath anymore...

But strangely enough, it built fine on debian's build machines on all
supported architectures:

  https://buildd.debian.org/status/package.php?p=guix&suite=experimental

*shrug*


live well,
  vagrant
[signature.asc (application/pgp-signature, inline)]

Information forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Sun, 02 May 2021 22:00:02 GMT) Full text and rfc822 format available.

Message #11 received at 48131 <at> debbugs.gnu.org (full text, mbox):

From: Ludovic Courtès <ludo <at> gnu.org>
To: Vagrant Cascadian <vagrant <at> debian.org>
Cc: 48131 <at> debbugs.gnu.org
Subject: Re: bug#48131: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Sun, 02 May 2021 23:59:47 +0200
Hi!

Vagrant Cascadian <vagrant <at> debian.org> skribis:

> On 2021-04-30, Vagrant Cascadian wrote:
>> I've been unable to build version 1.3.0rc1 for i386 on Debian:
>>
>>   [ 70%] GUILEC   gnu/packages/direct-connect.go
>>   [ 70%] GUILEC   gnu/packages/disk.go
>>   GC Warning: Failed to expand heap by 8388608 bytes
>>   GC Warning: Failed to expand heap by 8388608 bytes
>>   ...
>>   GC Warning: Failed to expand heap by 8388608 bytes
>>   GC Warning: Failed to expand heap by 8388608 bytes
>>   Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS
>>   make[3]: *** [Makefile:6871: make-go] Aborted
>>
>>
>> Debian is still using guile-2.2, so I'm guessing this is not a
>> particularly well tested codepath anymore...
>
> But strangely enough, it built fine on debian's build machines on all
> supported architectures:
>
>   https://buildd.debian.org/status/package.php?p=guix&suite=experimental

Guile 3.0’s compiler with ‘-O1’ (which is what we use for package files)
uses much less memory than 2.2¹.  The amount of memory may also be a
function of the number of threads used (‘-j’).  So it could be that the
build failed on a machine with less memory or more cores.

Ludo’.

¹ https://wingolog.org/archives/2020/06/03/a-baseline-compiler-for-guile




Information forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Sun, 02 May 2021 22:18:01 GMT) Full text and rfc822 format available.

Message #14 received at 48131 <at> debbugs.gnu.org (full text, mbox):

From: Vagrant Cascadian <vagrant <at> debian.org>
To: Ludovic Courtès <ludo <at> gnu.org>
Cc: 48131 <at> debbugs.gnu.org
Subject: Re: bug#48131: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Sun, 02 May 2021 15:17:47 -0700
[Message part 1 (text/plain, inline)]
On 2021-05-02, Ludovic Courtès wrote:
> Vagrant Cascadian <vagrant <at> debian.org> skribis:
>
>> On 2021-04-30, Vagrant Cascadian wrote:
>>> I've been unable to build version 1.3.0rc1 for i386 on Debian:
>>>
>>>   [ 70%] GUILEC   gnu/packages/direct-connect.go
>>>   [ 70%] GUILEC   gnu/packages/disk.go
>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>   ...
>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>   Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS
>>>   make[3]: *** [Makefile:6871: make-go] Aborted
>>>
>>>
>>> Debian is still using guile-2.2, so I'm guessing this is not a
>>> particularly well tested codepath anymore...
>>
>> But strangely enough, it built fine on debian's build machines on all
>> supported architectures:
>>
>>   https://buildd.debian.org/status/package.php?p=guix&suite=experimental
>
> Guile 3.0’s compiler with ‘-O1’ (which is what we use for package files)
> uses much less memory than 2.2¹.  The amount of memory may also be a
> function of the number of threads used (‘-j’).  So it could be that the
> build failed on a machine with less memory or more cores.

The machine it failed on had 32 cores and 140GB of ram... but will
experiment with reducing parallelism on that machine and see if that
helps.

The buildd.debian.org machine used only four cores, not sure how much
ram it had.


live well,
  vagrant
[signature.asc (application/pgp-signature, inline)]

Information forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Sun, 02 May 2021 22:51:02 GMT) Full text and rfc822 format available.

Message #17 received at 48131 <at> debbugs.gnu.org (full text, mbox):

From: Vagrant Cascadian <vagrant <at> debian.org>
To: Ludovic Courtès <ludo <at> gnu.org>
Cc: 48131 <at> debbugs.gnu.org
Subject: Re: bug#48131: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Sun, 02 May 2021 15:50:29 -0700
[Message part 1 (text/plain, inline)]
On 2021-05-02, Vagrant Cascadian wrote:
> On 2021-05-02, Ludovic Courtès wrote:
>> Vagrant Cascadian <vagrant <at> debian.org> skribis:
>>> On 2021-04-30, Vagrant Cascadian wrote:
>>>> I've been unable to build version 1.3.0rc1 for i386 on Debian:
>>>>
>>>>   [ 70%] GUILEC   gnu/packages/direct-connect.go
>>>>   [ 70%] GUILEC   gnu/packages/disk.go
>>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>>   ...
>>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>>   GC Warning: Failed to expand heap by 8388608 bytes
>>>>   Too many heap sections: Increase MAXHINCR or MAX_HEAP_SECTS
>>>>   make[3]: *** [Makefile:6871: make-go] Aborted
>>>>
>>>>
>>>> Debian is still using guile-2.2, so I'm guessing this is not a
>>>> particularly well tested codepath anymore...
>>>
>>> But strangely enough, it built fine on debian's build machines on all
>>> supported architectures:
>>>
>>>   https://buildd.debian.org/status/package.php?p=guix&suite=experimental
>>
>> Guile 3.0’s compiler with ‘-O1’ (which is what we use for package files)
>> uses much less memory than 2.2¹.  The amount of memory may also be a
>> function of the number of threads used (‘-j’).  So it could be that the
>> build failed on a machine with less memory or more cores.
>
> The machine it failed on had 32 cores and 140GB of ram... but will
> experiment with reducing parallelism on that machine and see if that
> helps.

I just realized that the Debian guix packages are built with parallelism
disabled for reproducible builds, so that seems unlikely to be the
issue... it might just be an issue with guile-2.2 on Debian...


live well,
  vagrant
[signature.asc (application/pgp-signature, inline)]

Information forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Mon, 03 May 2021 19:56:01 GMT) Full text and rfc822 format available.

Message #20 received at 48131 <at> debbugs.gnu.org (full text, mbox):

From: Ludovic Courtès <ludo <at> gnu.org>
To: Vagrant Cascadian <vagrant <at> debian.org>
Cc: 48131 <at> debbugs.gnu.org
Subject: Re: bug#48131: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Mon, 03 May 2021 21:54:41 +0200
Vagrant Cascadian <vagrant <at> debian.org> skribis:

> On 2021-05-02, Vagrant Cascadian wrote:
>> On 2021-05-02, Ludovic Courtès wrote:

[...]

>>> Guile 3.0’s compiler with ‘-O1’ (which is what we use for package files)
>>> uses much less memory than 2.2¹.  The amount of memory may also be a
>>> function of the number of threads used (‘-j’).  So it could be that the
>>> build failed on a machine with less memory or more cores.
>>
>> The machine it failed on had 32 cores and 140GB of ram... but will
>> experiment with reducing parallelism on that machine and see if that
>> helps.
>
> I just realized that the Debian guix packages are built with parallelism
> disabled for reproducible builds, so that seems unlikely to be the
> issue... it might just be an issue with guile-2.2 on Debian...

D’oh.  Could you try reducing the number of libgc marker thread, in case
that is what’s causing troubles?  You can do that by setting the
GC_MARKERS environment variables, as in GC_MARKERS=2.

Thanks,
Ludo’.




Information forwarded to bug-guix <at> gnu.org:
bug#48131; Package guix. (Fri, 18 Mar 2022 02:32:01 GMT) Full text and rfc822 format available.

Message #23 received at 48131 <at> debbugs.gnu.org (full text, mbox):

From: Maxim Cournoyer <maxim.cournoyer <at> gmail.com>
To: Ludovic Courtès <ludo <at> gnu.org>
Cc: Vagrant Cascadian <vagrant <at> debian.org>, 48131 <at> debbugs.gnu.org
Subject: Re: bug#48131: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Thu, 17 Mar 2022 22:31:31 -0400
Hi Vagrant,

Ludovic Courtès <ludo <at> gnu.org> writes:

> Vagrant Cascadian <vagrant <at> debian.org> skribis:
>
>> On 2021-05-02, Vagrant Cascadian wrote:
>>> On 2021-05-02, Ludovic Courtès wrote:
>
> [...]
>
>>>> Guile 3.0’s compiler with ‘-O1’ (which is what we use for package files)
>>>> uses much less memory than 2.2¹.  The amount of memory may also be a
>>>> function of the number of threads used (‘-j’).  So it could be that the
>>>> build failed on a machine with less memory or more cores.
>>>
>>> The machine it failed on had 32 cores and 140GB of ram... but will
>>> experiment with reducing parallelism on that machine and see if that
>>> helps.
>>
>> I just realized that the Debian guix packages are built with parallelism
>> disabled for reproducible builds, so that seems unlikely to be the
>> issue... it might just be an issue with guile-2.2 on Debian...
>
> D’oh.  Could you try reducing the number of libgc marker thread, in case
> that is what’s causing troubles?  You can do that by setting the
> GC_MARKERS environment variables, as in GC_MARKERS=2.

Is this issue still actual?

If so, could you try what Ludovic suggested?  Otherwise, let's close it.

Thank you,

Maxim




Reply sent to Vagrant Cascadian <vagrant <at> debian.org>:
You have taken responsibility. (Fri, 18 Mar 2022 06:10:02 GMT) Full text and rfc822 format available.

Notification sent to Vagrant Cascadian <vagrant <at> debian.org>:
bug acknowledged by developer. (Fri, 18 Mar 2022 06:10:02 GMT) Full text and rfc822 format available.

Message #28 received at 48131-done <at> debbugs.gnu.org (full text, mbox):

From: Vagrant Cascadian <vagrant <at> debian.org>
To: Maxim Cournoyer <maxim.cournoyer <at> gmail.com>, Ludovic Courtès <ludo <at> gnu.org>
Cc: 48131-done <at> debbugs.gnu.org
Subject: Re: bug#48131: build failure: 1.3.0rc1 on Debian i386 with guile-2.2
Date: Thu, 17 Mar 2022 23:09:40 -0700
[Message part 1 (text/plain, inline)]
On 2022-03-17, Maxim Cournoyer wrote:
> Ludovic Courtès <ludo <at> gnu.org> writes:
>> Vagrant Cascadian <vagrant <at> debian.org> skribis:
>>> On 2021-05-02, Vagrant Cascadian wrote:
>>>> On 2021-05-02, Ludovic Courtès wrote:
>>>>> Guile 3.0’s compiler with ‘-O1’ (which is what we use for package files)
>>>>> uses much less memory than 2.2¹.  The amount of memory may also be a
>>>>> function of the number of threads used (‘-j’).  So it could be that the
>>>>> build failed on a machine with less memory or more cores.
>>>>
>>>> The machine it failed on had 32 cores and 140GB of ram... but will
>>>> experiment with reducing parallelism on that machine and see if that
>>>> helps.
>>>
>>> I just realized that the Debian guix packages are built with parallelism
>>> disabled for reproducible builds, so that seems unlikely to be the
>>> issue... it might just be an issue with guile-2.2 on Debian...
>>
>> D’oh.  Could you try reducing the number of libgc marker thread, in case
>> that is what’s causing troubles?  You can do that by setting the
>> GC_MARKERS environment variables, as in GC_MARKERS=2.
>
> Is this issue still actual?
>
> If so, could you try what Ludovic suggested?  Otherwise, let's close it.

The Debian packages have since switched to use guile 3.0... I'll close
it for now.

Can always reopen if it starts happening again.

live well,
  vagrant
[signature.asc (application/pgp-signature, inline)]

bug archived. Request was from Debbugs Internal Request <help-debbugs <at> gnu.org> to internal_control <at> debbugs.gnu.org. (Fri, 15 Apr 2022 11:24:06 GMT) Full text and rfc822 format available.

This bug report was last modified 1 year and 348 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.