All of lore.kernel.org
 help / color / mirror / Atom feed
From: "Alex Bennée" <alex.bennee@linaro.org>
To: "Philippe Mathieu-Daudé" <f4bug@amsat.org>
Cc: "Fam Zheng" <fam@euphon.net>,
	"Christian Borntraeger" <borntraeger@de.ibm.com>,
	"Philippe Mathieu-Daudé" <philmd@redhat.com>,
	qemu-devel@nongnu.org, "Salvatore Bonaccorso" <carnil@debian.org>
Subject: Re: [PATCH] tests/docker: Kludge <linux/swab.h> breakage by pinning linux-libc-dev
Date: Fri, 15 May 2020 13:06:29 +0100	[thread overview]
Message-ID: <87blmpcshm.fsf@linaro.org> (raw)
In-Reply-To: <2839ba1a-f4c4-3a4d-cabe-d6e0e1b6accc@amsat.org>


Philippe Mathieu-Daudé <f4bug@amsat.org> writes:

> On 5/13/20 5:15 PM, Alex Bennée wrote:
>> Philippe Mathieu-Daudé <f4bug@amsat.org> writes:
>> 
>>> Linux kernel commit d5767057c9a [1] aimed to fix an issue with the
>>> swab() declaration, but doing so it introduced the BITS_PER_LONG
>>> definition, without using the kernel __ prefix, leading to odd
>>> failures in userland code using kernel uapi headers, such:
>>>
>>>        CC      block/file-posix.o
>>>      In file included from include/qemu/timer.h:4,
>>>                       from include/qemu/timed-average.h:29,
>>>                       from include/block/accounting.h:28,
>>>                       from include/block/block_int.h:27,
>>>                       from block/file-posix.c:30:
>>>      /usr/include/linux/swab.h: In function `__swab':
>>>      include/qemu/bitops.h:20:34: error: "sizeof" is not defined, evaluates to 0 [-Werror=undef]
>>>         20 | #define BITS_PER_LONG           (sizeof (unsigned long) * BITS_PER_BYTE)
>>>            |                                  ^~~~~~
>>>      include/qemu/bitops.h:20:41: error: missing binary operator before token "("
>>>         20 | #define BITS_PER_LONG           (sizeof (unsigned long) * BITS_PER_BYTE)
>>>            |                                         ^
>>>      cc1: all warnings being treated as errors
>>>      make: *** [rules.mak:69: block/file-posix.o] Error 1
>>>
>>> The issue has been fixed in Linux kernel commit 467d12f5c78 [2].
>>> Some distributions already backported the first commit, but still
>>> lack the second one.
>>>
>>> This is the case for Debian Buster.
>>> The linux-libc-dev package contains the Linux kernel headers.
>>> Kernel commit d5767057c9a has been backported in package
>>> commit 1fb0eb7956 "Update to 4.19.116" [3], see debian/changelog:
>>>
>>>    +linux (4.19.116-1) UNRELEASED; urgency=medium
>>>    ...
>>>    +    - uapi: rename ext2_swab() to swab() and share globally in swab.h
>>>
>>> The previous released update before it is debian/4.19.98-1,
>>> released as debian/4.19.98-1+deb10u1.
>>>
>>> We can find this package in the Debian snapshot archives,
>>> luckily archived on 2020-04-28 21:20:54 (see [4]).
>>>
>>> QEMU use Debian based Docker images for most of its cross-builds,
>>> which are tested by our Shippable CI jobs. The current broken
>>> package makes most of our CI red. We can kludge this by using the
>>> latest package released before the breakage. Do so by pinning
>>> the package version (apt hold), and using the snapshot archives,
>>> similar to commit b4048a7cd1.
>>> We'll revert once the fix is backported on Debian.
>> May as well mention:
>>    https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=960271
>
> Sure (I remembered you talked about it on IRC, but was lazy to dig).
>
>> 
>>>
>>> Reference to commits:
>>> [1] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=d5767057c9a
>>> [2] https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=467d12f5c78
>>> [3] https://salsa.debian.org/kernel-team/linux/-/commit/1fb0eb7956
>>> [4] http://snapshot.debian.org/package/linux/4.19.98-1%2Bdeb10u1/#linux-libc-dev_4.19.98-1:2b:deb10u1
>>>
>>> Cc: Salvatore Bonaccorso <carnil@debian.org>
>>> Cc: Christian Borntraeger <borntraeger@de.ibm.com>
>>> Signed-off-by: Philippe Mathieu-Daudé <f4bug@amsat.org>
>>> ---
>>> Shippable green again:
>>> https://app.shippable.com/github/philmd/qemu/runs/759/summary/console
>>> ---
>>>   tests/docker/dockerfiles/debian10.docker | 16 ++++++++++++++++
>>>   1 file changed, 16 insertions(+)
>>>
>>> diff --git a/tests/docker/dockerfiles/debian10.docker b/tests/docker/dockerfiles/debian10.docker
>>> index 0769700a41..d7084ac283 100644
>>> --- a/tests/docker/dockerfiles/debian10.docker
>>> +++ b/tests/docker/dockerfiles/debian10.docker
>>> @@ -9,6 +9,15 @@
>>>   #
>>>   FROM debian:buster-slim
>>>   + # Use a snapshot known to work (see
>>> http://snapshot.debian.org/#Usage)
>>> +ENV DEBIAN_SNAPSHOT_DATE "20200428T212054Z"
>>> +RUN sed -i \
>>> +    "s%^deb \(https\?://\)deb.debian.org/debian/\? \(.*\)%deb [check-valid-until=no] \1snapshot.debian.org/archive/debian/${DEBIAN_SNAPSHOT_DATE} \2%" \
>>> +    /etc/apt/sources.list
>>> +RUN sed -i \
>>> +    "s%^deb \(https\?://\)security.debian.org/debian-security/\? \(.*\)%deb [check-valid-until=no] \1snapshot.debian.org/archive/debian-security/${DEBIAN_SNAPSHOT_DATE} \2%" \
>>> +    /etc/apt/sources.list
>>> +
>>>   # Duplicate deb line as deb-src
>>>   RUN cat /etc/apt/sources.list | sed "s/^deb\ /deb-src /" >> /etc/apt/sources.list
>>>   @@ -35,4 +44,11 @@ RUN apt update && \
>>>           texinfo \
>>>           $(apt-get -s build-dep qemu | egrep ^Inst | fgrep '[all]' | cut -d\  -f2)
>>>   +# Install common build utilities
>>> +RUN apt update && \
>>> +    DEBIAN_FRONTEND=noninteractive eatmydata \
>>> +    apt install -y --allow-downgrades \
>>> +        linux-libc-dev=4.19.98-1+deb10u1 && \
>>> +    apt-mark hold linux-libc-dev
>>> +
>> Do we need this bit? Surely if we have a snapshot URL we don't need
>> to
>> hold the package as it will never get updated?
>
> You are right, it is a remnant from the first approach (pin it without
> using snapshot) which didn't work. I'll remove this line and the
> comment in description.

Hmm it seems this is needed - without it I see other failures:

  https://app.shippable.com/github/stsquad/qemu/runs/1377/8/console

-- 
Alex Bennée


  reply	other threads:[~2020-05-15 12:07 UTC|newest]

Thread overview: 7+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2020-05-13 12:01 [PATCH] tests/docker: Kludge <linux/swab.h> breakage by pinning linux-libc-dev Philippe Mathieu-Daudé
2020-05-13 15:15 ` Alex Bennée
2020-05-13 15:26   ` Philippe Mathieu-Daudé
2020-05-15 12:06     ` Alex Bennée [this message]
2020-05-15 12:41       ` Philippe Mathieu-Daudé
2020-05-13 17:32   ` Salvatore Bonaccorso
2020-05-13 17:53     ` Alex Bennée

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=87blmpcshm.fsf@linaro.org \
    --to=alex.bennee@linaro.org \
    --cc=borntraeger@de.ibm.com \
    --cc=carnil@debian.org \
    --cc=f4bug@amsat.org \
    --cc=fam@euphon.net \
    --cc=philmd@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.