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: <1da78077-a050-a391-a002-9e25eabc89d1@csgroup.eu>
Date:   Sat, 27 Aug 2022 18:32:42 +0000
From:   Christophe Leroy <christophe.leroy@...roup.eu>
To:     Pali Rohár <pali@...nel.org>
CC:     Michael Ellerman <mpe@...erman.id.au>,
        Joel Stanley <joel@....id.au>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Paul Mackerras <paulus@...ba.org>,
        "linuxppc-dev@...ts.ozlabs.org" <linuxppc-dev@...ts.ozlabs.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2] powerpc/boot: Don't always pass -mcpu=powerpc when
 building 32-bit uImage



Le 27/08/2022 à 19:36, Pali Rohár a écrit :
> On Saturday 27 August 2022 17:31:10 Christophe Leroy wrote:
>> Le 27/08/2022 à 15:39, Pali Rohár a écrit :
>>> For 32-bit uImage try to use CONFIG_TARGET_CPU option for -mcpu. This fixes
>>> following compiler error when building kernel with powerpc e500 SPE capable
>>> cross compilers:
>>>
>>>       BOOTAS  arch/powerpc/boot/crt0.o
>>>     powerpc-linux-gnuspe-gcc: error: unrecognized argument in option ‘-mcpu=powerpc’
>>>     powerpc-linux-gnuspe-gcc: note: valid arguments to ‘-mcpu=’ are: 8540 8548 native
>>>     make[1]: *** [arch/powerpc/boot/Makefile:231: arch/powerpc/boot/crt0.o] Error 1
>>>
>>> For 64-bit uImage and 64-bit kernels with 32-bit uImage wrapper there is no
>>> change.
>>>
>>> Similar change was already introduced for the main powerpc Makefile in
>>> commit 446cda1b21d9 ("powerpc/32: Don't always pass -mcpu=powerpc to the
>>> compiler").
>>>
>>> Fixes: 40a75584e526 ("powerpc/boot: Build wrapper for an appropriate CPU")
>>> Cc: stable@...r.kernel.org
>>> Signed-off-by: Pali Rohár <pali@...nel.org>
>>> ---
>>>    arch/powerpc/boot/Makefile | 8 ++++++++
>>>    1 file changed, 8 insertions(+)
>>>
>>> diff --git a/arch/powerpc/boot/Makefile b/arch/powerpc/boot/Makefile
>>> index a9cd2ea4a861..f56a5f90a5d8 100644
>>> --- a/arch/powerpc/boot/Makefile
>>> +++ b/arch/powerpc/boot/Makefile
>>> @@ -44,8 +44,16 @@ else
>>>    BOOTCFLAGS	+= -m64 -mcpu=powerpc64
>>>    endif
>>>    else
>>> +ifdef CONFIG_PPC32
>>> +ifdef CONFIG_TARGET_CPU_BOOL
>>> +BOOTCFLAGS	+= -m32 -mcpu=$(CONFIG_TARGET_CPU)
>>> +else
>>> +BOOTCFLAGS	+= -m32 -mcpu=powerpc
>>
>> You can't do that. You get here only if user has selected
>> TOOLCHAIN_DEFAULT_CPU, in which case you don't want to for -mcpu=powerpc.
> 
> So do I understand it correctly that in this branch I should omit -mcpu=powerpc?

Correct.

> 
>> -mcpu=powerpc is set when user selects CONFIG_POWERPC_CPU, in which case
>> CONFIG_TARGET_CPU_BOOL is set as well.
>>
>>> +endif
>>> +else
>>>    BOOTCFLAGS	+= -m32 -mcpu=powerpc
>>
>> Same, for PPC64 I think you don't want that either, unless the
>> CONFIG_GENERIC_CPU has been selected in which case
>> CONFIG_TARGET_CPU_BOOL is not set.
>>
>> When CONFIG_TARGET_CPU_BOOL is set for PPC64 you also want
>> -mcpu=$(CONFIG_TARGET_CPU)
> 
> I understand that this branch is called for PPC64 build with 32-bit
> uImage wrapper. So in this case should not be used TARGET_CPU as it
> would be 64-bit and not 32-bit as requited for 32-bit uImage wrapper.

Why ?

-mcpu=e6500 -m32 works as far as I can see.

> 
> Anyway, in this change I'm touching only PPC32 build, so all PPC64 stay
> as it was before.
> 
>>>    endif
>>> +endif
>>>    
>>>    BOOTCFLAGS	+= -isystem $(shell $(BOOTCC) -print-file-name=include)
>>>    

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ