linux-kernel - Re: [PATCH 00/16] mm: prepare for converting vm->vm

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]

Date:	Wed, 21 Mar 2012 17:16:06 +0400
From:	Konstantin Khlebnikov <khlebnikov@...nvz.org>
To:	Minchan Kim <minchan@...nel.org>
CC:	Andrew Morton <akpm@...ux-foundation.org>,
	"linux-mm@...ck.org" <linux-mm@...ck.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Linus Torvalds <torvalds@...ux-foundation.org>,
	Hugh Dickins <hughd@...gle.com>,
	KOSAKI Motohiro <kosaki.motohiro@...fujitsu.com>,
	Ben Herrenschmidt <benh@...nel.crashing.org>,
	"linux@....linux.org.uk" <linux@....linux.org.uk>
Subject: Re: [PATCH 00/16] mm: prepare for converting vm->vm_flags to 64-bit

Minchan Kim wrote:
> Hi Konstantin,
>
> It seems to be nice clean up to me and you are a volunteer we have been wanted
> for a long time. Thanks!
> I am one of people who really want to expand vm_flags to 64 bit but when KOSAKI
> tried it, Linus said his concerning, I guess you already saw that.
>
> He want to tidy vm_flags's usage up rather than expanding it.
> Without the discussion about that, just expanding vm_flags would make us use
> it up easily so that we might need more space.

Strictly speaking, my pachset does not expands vm_flags, it just prepares to this.
Anyway vm_flags_t looks better than hard-coded "unsigned long" and messy type-casts around it.

>
> Readahead flags are good candidate to move into another space and arch-specific flags, I guess.
> Another candidate I think of is THP flag. It's just for only anonymous vma now
> (But I am not sure we have a plan to support it for file-backed pages in future)
> so we can move it to anon_vma or somewhere.
> I think other guys might find more somethings
>
> The point is that at least, we have to discuss about clean up current vm_flags's
> use cases before expanding it unconditionally.

Seems like we can easily remove VM_EXECUTABLE
(count in mm->num_exe_file_vmas amount of vmas with vma->vm_file == mm->exe_file
instead of vmas with VM_EXECUTABLE bit)

And probably VM_CAN_NONLINEAR...

>
> On Wed, Mar 21, 2012 at 10:56:07AM +0400, Konstantin Khlebnikov wrote:
>> There is good old tradition: every year somebody submit patches for extending
>> vma->vm_flags upto 64-bits, because there no free bits left on 32-bit systems.
>>
>> previous attempts:
>> https://lkml.org/lkml/2011/4/12/24	(KOSAKI Motohiro)
>> https://lkml.org/lkml/2010/4/27/23	(Benjamin Herrenschmidt)
>> https://lkml.org/lkml/2009/10/1/202	(Hugh Dickins)
>>
>> Here already exist special type for this: vm_flags_t, but not all code uses it.
>> So, before switching vm_flags_t from unsinged long to u64 we must spread
>> vm_flags_t everywhere and fix all possible type-casting problems.
>>
>> There is no functional changes in this patch set,
>> it only prepares code for vma->vm_flags converting.
>>
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/