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] [day] [month] [year] [list]
Date:   Wed, 25 Oct 2023 16:15:31 +0200
From:   Thomas Huth <>
To:     Arnd Bergmann <>, Oleg Nesterov <>,, Brian Cain <>
Cc:, Richard Kuo <>
Subject: Re: [PATCH] hexagon: Remove unusable symbols from the ptrace.h uapi

On 25/10/2023 15.59, Arnd Bergmann wrote:
> On Wed, Oct 25, 2023, at 09:38, Thomas Huth wrote:
>> Kernel-internal prototypes, references to current_thread_info()
>> and code hidden behind a CONFIG_HEXAGON_ARCH_VERSION switch are
>> certainly not usable in userspace, so this should not reside
>> in a uapi header. Move the code into an internal version of
>> ptrace.h instead.
>> Signed-off-by: Thomas Huth <>
>> ---
>>   I've compile tested it now with a hexagon cross-compiler and the kernel
>>   compiles fine with this change, so I think this should be good to go.
> I've applied this to the asm-generic tree, thanks for the
> patch.


>> +++ b/scripts/
>> @@ -74,7 +74,6 @@ arch/arc/include/uapi/asm/page.h:CONFIG_ARC_PAGE_SIZE_16K
>>   arch/arc/include/uapi/asm/page.h:CONFIG_ARC_PAGE_SIZE_4K
>>   arch/arc/include/uapi/asm/swab.h:CONFIG_ARC_HAS_SWAPE
>>   arch/arm/include/uapi/asm/ptrace.h:CONFIG_CPU_ENDIAN_BE8
>> -arch/hexagon/include/uapi/asm/ptrace.h:CONFIG_HEXAGON_ARCH_VERSION
>>   arch/hexagon/include/uapi/asm/user.h:CONFIG_HEXAGON_ARCH_VERSION
>>   arch/m68k/include/uapi/asm/ptrace.h:CONFIG_COLDFIRE
>>   arch/nios2/include/uapi/asm/swab.h:CONFIG_NIOS2_CI_SWAB_NO
> Would you like to send another patch for the other hexagon
> file? It looks trivial enough as we can just drop the #if
> portion there and keep the #else side.

Looks like we have at least to look carefully at 
arch/hexagon/kernel/ptrace.c first ... pad1 is still used there and the of 
offsetof(struct user_regs_struct, pad1) results in different values 
depending on the CONFIG switch ... but sure, I can have a try to come up 
with a patch.


Powered by blists - more mailing lists