Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > linux.kernel > #1306995 > unrolled thread

linux-next: build failure after merge of the aio tree

Started byStephen Rothwell <sfr@canb.auug.org.au>
First post2016-01-12 06:50 +0100
Last post2016-01-15 10:30 +0100
Articles 9 — 3 participants

Back to article view | Back to linux.kernel

Contents

  linux-next: build failure after merge of the aio tree Stephen Rothwell <sfr@canb.auug.org.au> - 2016-01-12 06:50 +0100
    Re: linux-next: build failure after merge of the aio tree Benjamin LaHaise <bcrl@kvack.org> - 2016-01-12 17:40 +0100
      Re: linux-next: build failure after merge of the aio tree Stephen Rothwell <sfr@canb.auug.org.au> - 2016-01-27 03:50 +0100
    Re: linux-next: build failure after merge of the aio tree Stephen Rothwell <sfr@canb.auug.org.au> - 2016-01-15 03:30 +0100
    Re: linux-next: build failure after merge of the aio tree Christoph Hellwig <hch@infradead.org> - 2016-01-15 08:40 +0100
      Re: linux-next: build failure after merge of the aio tree Christoph Hellwig <hch@infradead.org> - 2016-01-15 10:30 +0100
        Re: linux-next: build failure after merge of the aio tree Benjamin LaHaise <bcrl@kvack.org> - 2016-01-15 16:20 +0100
          Re: linux-next: build failure after merge of the aio tree Stephen Rothwell <sfr@canb.auug.org.au> - 2016-01-16 00:00 +0100
      Re: linux-next: build failure after merge of the aio tree Stephen Rothwell <sfr@canb.auug.org.au> - 2016-01-15 10:30 +0100

#1306995 — linux-next: build failure after merge of the aio tree

FromStephen Rothwell <sfr@canb.auug.org.au>
Date2016-01-12 06:50 +0100
Subjectlinux-next: build failure after merge of the aio tree
Message-ID<qQ009-51J-1@gated-at.bofh.it>
Hi Benjamin,

After merging the aio tree, today's linux-next build (arm
multi_v7_defconfig) failed like this:

fs/built-in.o: In function `aio_thread_op_foo_at':
file.c:(.text+0x43808): undefined reference to `__get_user_bad'
file.c:(.text+0x43838): undefined reference to `__get_user_bad'

Caused by commit

  150a0b4905f1 ("aio: add support for async openat()")

I have used the aio tree from next-20160111 for today.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[toc] | [next] | [standalone]

#1307615

FromBenjamin LaHaise <bcrl@kvack.org>
Date2016-01-12 17:40 +0100
Message-ID<qQa9d-3tC-25@gated-at.bofh.it>
In reply to#1306995
On Tue, Jan 12, 2016 at 04:40:34PM +1100, Stephen Rothwell wrote:
> Hi Benjamin,
> 
> After merging the aio tree, today's linux-next build (arm
> multi_v7_defconfig) failed like this:
> 
> fs/built-in.o: In function `aio_thread_op_foo_at':
> file.c:(.text+0x43808): undefined reference to `__get_user_bad'
> file.c:(.text+0x43838): undefined reference to `__get_user_bad'

This is very strange.  It seems to imply that __get_user() doesn't 
handle 64 bit values, which is completely broken behaviour on the 
architecture's part if true.  Can any ARM folks comment on the right 
fix here?

		-ben

> Caused by commit
> 
>   150a0b4905f1 ("aio: add support for async openat()")
> 
> I have used the aio tree from next-20160111 for today.
> 
> -- 
> Cheers,
> Stephen Rothwell                    sfr@canb.auug.org.au

-- 
"Thought is the essence of where you are now."

[toc] | [prev] | [next] | [standalone]

#1318571

FromStephen Rothwell <sfr@canb.auug.org.au>
Date2016-01-27 03:50 +0100
Message-ID<qVolc-57c-1@gated-at.bofh.it>
In reply to#1307615
Hi Benjamin,

On Tue, 12 Jan 2016 11:38:35 -0500 Benjamin LaHaise <bcrl@kvack.org> wrote:
>
> On Tue, Jan 12, 2016 at 04:40:34PM +1100, Stephen Rothwell wrote:
> > 
> > After merging the aio tree, today's linux-next build (arm
> > multi_v7_defconfig) failed like this:
> > 
> > fs/built-in.o: In function `aio_thread_op_foo_at':
> > file.c:(.text+0x43808): undefined reference to `__get_user_bad'
> > file.c:(.text+0x43838): undefined reference to `__get_user_bad'  
> 
> This is very strange.  It seems to imply that __get_user() doesn't 
> handle 64 bit values, which is completely broken behaviour on the 
> architecture's part if true.  Can any ARM folks comment on the right 
> fix here?

Well, probably only if you cc them :-)

Indeed, __get_user on arm does not handle 64 bit objects, where as
get_user does ...

Background: new aio code is adding __get_user() calls referencing 64
bit quantities (__u64 and __s64).

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[toc] | [prev] | [next] | [standalone]

#1309823

FromStephen Rothwell <sfr@canb.auug.org.au>
Date2016-01-15 03:30 +0100
Message-ID<qR2jf-7U0-9@gated-at.bofh.it>
In reply to#1306995
Hi Benjamin,

On Tue, 12 Jan 2016 16:40:34 +1100 Stephen Rothwell <sfr@canb.auug.org.au> wrote:
>
> After merging the aio tree, today's linux-next build (arm
> multi_v7_defconfig) failed like this:
> 
> fs/built-in.o: In function `aio_thread_op_foo_at':
> file.c:(.text+0x43808): undefined reference to `__get_user_bad'
> file.c:(.text+0x43838): undefined reference to `__get_user_bad'
> 
> Caused by commit
> 
>   150a0b4905f1 ("aio: add support for async openat()")
> 
> I have used the aio tree from next-20160111 for today.

I am still getting this error.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[toc] | [prev] | [next] | [standalone]

#1309928

FromChristoph Hellwig <hch@infradead.org>
Date2016-01-15 08:40 +0100
Message-ID<qR79f-2OU-3@gated-at.bofh.it>
In reply to#1306995
On Tue, Jan 12, 2016 at 04:40:34PM +1100, Stephen Rothwell wrote:
> Hi Benjamin,
> 
> After merging the aio tree, today's linux-next build (arm
> multi_v7_defconfig) failed like this:
> 
> fs/built-in.o: In function `aio_thread_op_foo_at':
> file.c:(.text+0x43808): undefined reference to `__get_user_bad'
> file.c:(.text+0x43838): undefined reference to `__get_user_bad'
> 
> Caused by commit
> 
>   150a0b4905f1 ("aio: add support for async openat()")

How did that code end up in linux-next anyway?

[toc] | [prev] | [next] | [standalone]

#1309985

FromChristoph Hellwig <hch@infradead.org>
Date2016-01-15 10:30 +0100
Message-ID<qR8RJ-464-21@gated-at.bofh.it>
In reply to#1309928
On Fri, Jan 15, 2016 at 08:23:16PM +1100, Stephen Rothwell wrote:
> Via the aio tree (git://git.kvack.org/~bcrl/aio-next.git#master) added
> in July 2013 at Ben's request.  The code was added to the aio tree in
> Jan 12 (my time), but has never been in a published linux-next tree due
> to the above build problem (I back out to the previous days version of
> the aio tree).

Well, it's code Ben posted a few days ago, which to say it mildly is
rather controversial.  It's cetainly not 4.5 material.

> 
> -- 
> Cheers,
> Stephen Rothwell                    sfr@canb.auug.org.au
---end quoted text---

[toc] | [prev] | [next] | [standalone]

#1310212

FromBenjamin LaHaise <bcrl@kvack.org>
Date2016-01-15 16:20 +0100
Message-ID<qRekp-7Xi-1@gated-at.bofh.it>
In reply to#1309985
On Fri, Jan 15, 2016 at 01:25:31AM -0800, Christoph Hellwig wrote:
> On Fri, Jan 15, 2016 at 08:23:16PM +1100, Stephen Rothwell wrote:
> > Via the aio tree (git://git.kvack.org/~bcrl/aio-next.git#master) added
> > in July 2013 at Ben's request.  The code was added to the aio tree in
> > Jan 12 (my time), but has never been in a published linux-next tree due
> > to the above build problem (I back out to the previous days version of
> > the aio tree).
> 
> Well, it's code Ben posted a few days ago, which to say it mildly is
> rather controversial.  It's cetainly not 4.5 material.

It still needs the exposure.

As for the build failure, it's a bug in the arch __get_user() implementation 
that needs to be fixed.  __get_user() should really be able to handle 64 bit 
types.

		-ben

> > -- 
> > Cheers,
> > Stephen Rothwell                    sfr@canb.auug.org.au
> ---end quoted text---

-- 
"Thought is the essence of where you are now."

[toc] | [prev] | [next] | [standalone]

#1310517

FromStephen Rothwell <sfr@canb.auug.org.au>
Date2016-01-16 00:00 +0100
Message-ID<qRlvz-46b-3@gated-at.bofh.it>
In reply to#1310212
Hi Ben,

On Fri, 15 Jan 2016 10:18:21 -0500 Benjamin LaHaise <bcrl@kvack.org> wrote:
>
> On Fri, Jan 15, 2016 at 01:25:31AM -0800, Christoph Hellwig wrote:
> > On Fri, Jan 15, 2016 at 08:23:16PM +1100, Stephen Rothwell wrote:  
> > > Via the aio tree (git://git.kvack.org/~bcrl/aio-next.git#master) added
> > > in July 2013 at Ben's request.  The code was added to the aio tree in
> > > Jan 12 (my time), but has never been in a published linux-next tree due
> > > to the above build problem (I back out to the previous days version of
> > > the aio tree).  
> > 
> > Well, it's code Ben posted a few days ago, which to say it mildly is
> > rather controversial.  It's cetainly not 4.5 material.  
> 
> It still needs the exposure.

If it is not destined for v4.5, then it should not (yet) be in
linux-next.  It should wait until after v4.5-rc1 is released (the merge
window closes).  I would also argue that if the functionality itself is
still under active review (and I haven't competely followed the
discussion so I don't know where that is up to, but Christoph, at
least, seems not completely convinced), then it should also not yet be
in linux-next.

> As for the build failure, it's a bug in the arch __get_user() implementation 
> that needs to be fixed.  __get_user() should really be able to handle 64 bit 
> types.

Yeah, it is a bit weird.

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[toc] | [prev] | [next] | [standalone]

#1309988

FromStephen Rothwell <sfr@canb.auug.org.au>
Date2016-01-15 10:30 +0100
Message-ID<qR8RJ-464-23@gated-at.bofh.it>
In reply to#1309928
Hi Christoph,

On Thu, 14 Jan 2016 23:39:44 -0800 Christoph Hellwig <hch@infradead.org> wrote:
>
> On Tue, Jan 12, 2016 at 04:40:34PM +1100, Stephen Rothwell wrote:
> > Hi Benjamin,
> > 
> > After merging the aio tree, today's linux-next build (arm
> > multi_v7_defconfig) failed like this:
> > 
> > fs/built-in.o: In function `aio_thread_op_foo_at':
> > file.c:(.text+0x43808): undefined reference to `__get_user_bad'
> > file.c:(.text+0x43838): undefined reference to `__get_user_bad'
> > 
> > Caused by commit
> > 
> >   150a0b4905f1 ("aio: add support for async openat()")  
> 
> How did that code end up in linux-next anyway?

Via the aio tree (git://git.kvack.org/~bcrl/aio-next.git#master) added
in July 2013 at Ben's request.  The code was added to the aio tree in
Jan 12 (my time), but has never been in a published linux-next tree due
to the above build problem (I back out to the previous days version of
the aio tree).

-- 
Cheers,
Stephen Rothwell                    sfr@canb.auug.org.au

[toc] | [prev] | [standalone]

Back to top | Article view | linux.kernel

csiph-web