linux-kernel - Re: [PATCH v3 06/15] powerpc/32: prepare for CONFIG_VMAP

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: <d3908c21-553a-029e-d671-2cd3628776dc@linux.ibm.com>
Date:   Fri, 18 Oct 2019 18:09:29 +1100
From:   Andrew Donnellan <ajd@...ux.ibm.com>
To:     Christophe Leroy <christophe.leroy@....fr>,
        Benjamin Herrenschmidt <benh@...nel.crashing.org>,
        Paul Mackerras <paulus@...ba.org>,
        Michael Ellerman <mpe@...erman.id.au>, npiggin@...il.com,
        dja@...ens.net
Cc:     linux-kernel@...r.kernel.org, linuxppc-dev@...ts.ozlabs.org,
        linux-mm@...ck.org
Subject: Re: [PATCH v3 06/15] powerpc/32: prepare for CONFIG_VMAP_STACK

On 18/10/19 5:46 pm, Christophe Leroy wrote:
> 
> 
> Le 17/10/2019 à 09:36, Andrew Donnellan a écrit :
>> On 10/9/19 7:16 pm, Christophe Leroy wrote:
>>> +#if defined(CONFIG_VMAP_STACK) && CONFIG_THREAD_SHIFT < PAGE_SHIFT
>>> +#define THREAD_SHIFT        PAGE_SHIFT
>>> +#else
>>>   #define THREAD_SHIFT        CONFIG_THREAD_SHIFT
>>> +#endif
>>>
>>>   #define THREAD_SIZE        (1 << THREAD_SHIFT)
>>>
>>
>> Looking at 64-bit book3s: with 64K pages, this results in a 
>> THREAD_SIZE that's too large for immediate mode arithmetic operations, 
>> which is annoying. Hmm.
>>
> 
> Which operation are you thinking about ?
> 
> For instance, 'addi' can't be used anymore, but 'addis' can.

Right, I didn't look too closely - yes, they're "addi"s. But also 
DS-form stores:

	stdu	r1,THREAD_SIZE-STACK_FRAME_OVERHEAD(r3)

Anyway, this is just the first thing that broke when I wondered whether 
a ppc64le_defconfig would even build if I just switched on VMAP_STACK, 
I'll work around it

-- 
Andrew Donnellan              OzLabs, ADL Canberra
ajd@...ux.ibm.com             IBM Australia Limited