linux-arch.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* Re: [PATCH] ANDROID: binder: remove 32-bit binder interface.
       [not found]     ` <20180514120338.GA5265@infradead.org>
@ 2018-05-14 14:00       ` Geert Uytterhoeven
  2018-05-14 14:00         ` Geert Uytterhoeven
  2018-05-14 21:30         ` Martijn Coenen
  0 siblings, 2 replies; 4+ messages in thread
From: Geert Uytterhoeven @ 2018-05-14 14:00 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: Martijn Coenen, kbuild test robot, Greg KH, kbuild-all,
	John Stultz, Todd Kjos, Arve Hjønnevåg, Amit Pundir,
	Arnd Bergmann, LKML, open list:ANDROID DRIVERS, Martijn Coenen,
	linux-m68k, Linux-Arch

Hi Christoph,

On Mon, May 14, 2018 at 2:03 PM, Christoph Hellwig <hch@infradead.org> wrote:
> On Fri, May 11, 2018 at 09:57:52AM +0200, Martijn Coenen wrote:
>> On Sat, May 5, 2018 at 2:10 PM, kbuild test robot <lkp@intel.com> wrote:
>> >    drivers/android/binder.o: In function `binder_thread_write':
>> >>> binder.c:(.text+0x6a16): undefined reference to `__get_user_bad'
>>
>> Looks like m68k doesn't support 64-bit get_user(). I could just have
>> binder depend on !CONFIG_M68K, but there may be other architectures
>> still that don't support this. Another alternative would be to
>> whitelist the architectures Android supports - eg arm, arm64, x86,
>> x86_64. But I'm not sure if arch-limited drivers are considered bad
>> form. Does anybody have suggestions for how to deal with this?
>
> The proper fix is to just support 640bit get/put_user on m68k instead

I hope we'll never need 640bit support in {get,put}_user() ;-)

> of working around this.

Patch sent.

BTW, sh also doesn't seem to have 64-bit get_user().
There may be others.

BTW2, does the Android Binder need to care about endianness when talking
to userspace?

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] ANDROID: binder: remove 32-bit binder interface.
  2018-05-14 14:00       ` [PATCH] ANDROID: binder: remove 32-bit binder interface Geert Uytterhoeven
@ 2018-05-14 14:00         ` Geert Uytterhoeven
  2018-05-14 21:30         ` Martijn Coenen
  1 sibling, 0 replies; 4+ messages in thread
From: Geert Uytterhoeven @ 2018-05-14 14:00 UTC (permalink / raw)
  To: Christoph Hellwig
  Cc: Martijn Coenen, kbuild test robot, Greg KH, kbuild-all,
	John Stultz, Todd Kjos, Arve Hjønnevåg, Amit Pundir,
	Arnd Bergmann, LKML, open list:ANDROID DRIVERS, Martijn Coenen,
	linux-m68k, Linux-Arch

Hi Christoph,

On Mon, May 14, 2018 at 2:03 PM, Christoph Hellwig <hch@infradead.org> wrote:
> On Fri, May 11, 2018 at 09:57:52AM +0200, Martijn Coenen wrote:
>> On Sat, May 5, 2018 at 2:10 PM, kbuild test robot <lkp@intel.com> wrote:
>> >    drivers/android/binder.o: In function `binder_thread_write':
>> >>> binder.c:(.text+0x6a16): undefined reference to `__get_user_bad'
>>
>> Looks like m68k doesn't support 64-bit get_user(). I could just have
>> binder depend on !CONFIG_M68K, but there may be other architectures
>> still that don't support this. Another alternative would be to
>> whitelist the architectures Android supports - eg arm, arm64, x86,
>> x86_64. But I'm not sure if arch-limited drivers are considered bad
>> form. Does anybody have suggestions for how to deal with this?
>
> The proper fix is to just support 640bit get/put_user on m68k instead

I hope we'll never need 640bit support in {get,put}_user() ;-)

> of working around this.

Patch sent.

BTW, sh also doesn't seem to have 64-bit get_user().
There may be others.

BTW2, does the Android Binder need to care about endianness when talking
to userspace?

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] ANDROID: binder: remove 32-bit binder interface.
  2018-05-14 14:00       ` [PATCH] ANDROID: binder: remove 32-bit binder interface Geert Uytterhoeven
  2018-05-14 14:00         ` Geert Uytterhoeven
@ 2018-05-14 21:30         ` Martijn Coenen
  2018-05-14 21:30           ` Martijn Coenen
  1 sibling, 1 reply; 4+ messages in thread
From: Martijn Coenen @ 2018-05-14 21:30 UTC (permalink / raw)
  To: Geert Uytterhoeven
  Cc: Amit Pundir, open list:ANDROID DRIVERS, Arnd Bergmann,
	linux-m68k, Greg KH, LKML, Christoph Hellwig,
	Arve Hjønnevåg, Martijn Coenen, John Stultz,
	kbuild-all, Linux-Arch, Todd Kjos

On Mon, May 14, 2018 at 4:00 PM, Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
> Patch sent.

Thanks for the quick turn-around!

>
> BTW, sh also doesn't seem to have 64-bit get_user().
> There may be others.

I checked quickly, nios2 is the only other arch that explicitly
doesn't support it and would result in a build error; some other archs
don't define __get_user, but in that case they just fall back to
raw_copy_from_user().

>
> BTW2, does the Android Binder need to care about endianness when talking
> to userspace?

No, I don't think it should.

Thanks,
Martijn

>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH] ANDROID: binder: remove 32-bit binder interface.
  2018-05-14 21:30         ` Martijn Coenen
@ 2018-05-14 21:30           ` Martijn Coenen
  0 siblings, 0 replies; 4+ messages in thread
From: Martijn Coenen @ 2018-05-14 21:30 UTC (permalink / raw)
  To: Geert Uytterhoeven
  Cc: Christoph Hellwig, kbuild test robot, Greg KH, kbuild-all,
	John Stultz, Todd Kjos, Arve Hjønnevåg, Amit Pundir,
	Arnd Bergmann, LKML, open list:ANDROID DRIVERS, Martijn Coenen,
	linux-m68k, Linux-Arch

On Mon, May 14, 2018 at 4:00 PM, Geert Uytterhoeven
<geert@linux-m68k.org> wrote:
> Patch sent.

Thanks for the quick turn-around!

>
> BTW, sh also doesn't seem to have 64-bit get_user().
> There may be others.

I checked quickly, nios2 is the only other arch that explicitly
doesn't support it and would result in a build error; some other archs
don't define __get_user, but in that case they just fall back to
raw_copy_from_user().

>
> BTW2, does the Android Binder need to care about endianness when talking
> to userspace?

No, I don't think it should.

Thanks,
Martijn

>
> Gr{oetje,eeting}s,
>
>                         Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2018-05-14 21:30 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
     [not found] <20180504195728.72932-1-maco@android.com>
     [not found] ` <201805060540.zjwonWcp%fengguang.wu@intel.com>
     [not found]   ` <CAB0TPYG-Q6Ki_FMM1u+i_g+0ykQgeUTRpm00qycAUVQOqVH6eQ@mail.gmail.com>
     [not found]     ` <20180514120338.GA5265@infradead.org>
2018-05-14 14:00       ` [PATCH] ANDROID: binder: remove 32-bit binder interface Geert Uytterhoeven
2018-05-14 14:00         ` Geert Uytterhoeven
2018-05-14 21:30         ` Martijn Coenen
2018-05-14 21:30           ` Martijn Coenen

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).