lists.openwall.net   lists  /  announce  owl-users  owl-dev  john-users  john-dev  passwdqc-users  yescrypt  popa3d-users  /  oss-security  kernel-hardening  musl  sabotage  tlsify  passwords  /  crypt-dev  xvendor  /  Bugtraq  Full-Disclosure  linux-kernel  linux-netdev  linux-ext4  linux-hardening  linux-cve-announce  PHC 
Open Source and information security mailing list archives
 
Hash Suite: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <325243a63755a1f6ae578dae35154b17@agner.ch>
Date:   Tue, 12 Jun 2018 19:27:29 +0200
From:   Stefan Agner <stefan@...er.ch>
To:     Guenter Roeck <linux@...ck-us.net>
Cc:     linux@...linux.org.uk, ard.biesheuvel@...aro.org, arnd@...db.de,
        nicolas.pitre@...aro.org, keescook@...omium.org,
        marc.zyngier@....com, linux-kernel@...r.kernel.org,
        mka@...omium.org, robin.murphy@....com,
        linux-arm-kernel@...ts.infradead.org,
        Bernhard.Rosenkranzer@...aro.org
Subject: Re: [v2,4/6] ARM: drop no-thumb-interwork in EABI mode

On 12.06.2018 19:19, Guenter Roeck wrote:
> On Sun, Mar 25, 2018 at 08:09:57PM +0200, Stefan Agner wrote:
>> According to GCC documentation -m(no-)thumb-interwork is
>> meaningless in AAPCS configurations. Also clang does not
> 
> It appears that this is only correct for recent versions of gcc.
> 
> With gcc 4.9.2, this patch causes the qemu collie emulation
> to fail with collie_defconfig+CONFIG_AEABI.

Hm, interesting. However, even 4.9.0 claims this option is meaningless
when using AAPCS configurations:
https://gcc.gnu.org/onlinedocs/gcc-4.9.0/gcc/ARM-Options.html#ARM-Options

> 
> qemu-system-arm: Trying to execute code outside RAM or ROM at 0x02000000
> This usually means one of the following happened:
> ...
> 
> With gcc 7.3.0, the emulation works as expected. Reverting the patch
> fixes the problem with gcc 4.9.2. Not that it matters much to me - I can
> and will switch to gcc 7.3.0 for my testing - but effectively this means
> that older versions of gcc are no longer supported for all configurations.
> 
> Maybe $(call cc-option,-mno-thumb-interwork,) would have been safer ?

I used to have call cc-option in place, but I removed that when I
realized that gcc claims it is meaningless with AAPCS configurations.

--
Stefan

> 
> Guenter
> 
>> support the flag:
>>   clang-5.0: error: unknown argument: '-mno-thumb-interwork'
>>
>> Just drop -mno-thumb-interwork in AEABI configuration.
>>
>> Signed-off-by: Stefan Agner <stefan@...er.ch>
>> ---
>>  arch/arm/Makefile | 2 +-
>>  1 file changed, 1 insertion(+), 1 deletion(-)
>>
>> diff --git a/arch/arm/Makefile b/arch/arm/Makefile
>> index e83f5161fdd8..e9e3fde3c657 100644
>> --- a/arch/arm/Makefile
>> +++ b/arch/arm/Makefile
>> @@ -106,7 +106,7 @@ tune-$(CONFIG_CPU_V6K)		=$(call cc-option,-mtune=arm1136j-s,-mtune=strongarm)
>>  tune-y := $(tune-y)
>>
>>  ifeq ($(CONFIG_AEABI),y)
>> -CFLAGS_ABI	:=-mabi=aapcs-linux -mno-thumb-interwork -mfpu=vfp
>> +CFLAGS_ABI	:=-mabi=aapcs-linux -mfpu=vfp
>>  else
>>  CFLAGS_ABI	:=$(call cc-option,-mapcs-32,-mabi=apcs-gnu) $(call cc-option,-mno-thumb-interwork,)
>>  endif

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ