linux-kernel - Re: [PATCH v9 bpf-next 1/9] x86/Kconfig: select HAVE_ARCH_HUGE_VMALLOC with HAVE_ARCH_HUGE

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: <e05d99f4b8b8719f99e1de44dc26e94c9994c34b.camel@intel.com>
Date:   Tue, 29 Mar 2022 21:36:09 +0000
From:   "Edgecombe, Rick P" <rick.p.edgecombe@...el.com>
To:     "songliubraving@...com" <songliubraving@...com>
CC:     "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "daniel@...earbox.net" <daniel@...earbox.net>,
        "peterz@...radead.org" <peterz@...radead.org>,
        "ast@...nel.org" <ast@...nel.org>,
        "bpf@...r.kernel.org" <bpf@...r.kernel.org>,
        "Kernel-team@...com" <Kernel-team@...com>,
        "song@...nel.org" <song@...nel.org>, "hch@....de" <hch@....de>,
        "pmenzel@...gen.mpg.de" <pmenzel@...gen.mpg.de>,
        "andrii@...nel.org" <andrii@...nel.org>,
        "x86@...nel.org" <x86@...nel.org>,
        "iii@...ux.ibm.com" <iii@...ux.ibm.com>,
        "netdev@...r.kernel.org" <netdev@...r.kernel.org>,
        "urezki@...il.com" <urezki@...il.com>,
        "npiggin@...il.com" <npiggin@...il.com>
Subject: Re: [PATCH v9 bpf-next 1/9] x86/Kconfig: select
 HAVE_ARCH_HUGE_VMALLOC with HAVE_ARCH_HUGE_VMAP

On Tue, 2022-03-29 at 19:13 +0000, Song Liu wrote:
> > 
> > I don't think we should tie this to vmap_allow_huge by definition.
> > Also, what it does is try 2MB pages for allocations larger than
> > 2MB.
> > For smaller allocations it doesn't try or "prefer" them.
> 
> How about something like:
> 
> #define VM_TRY_HUGE_VMAP        0x00001000      /* try PMD_SIZE
> mapping if size-per-node >= PMD_SIZE */

Seems reasonable name. I don't know if "size-per-node >= PMD_SIZE" is
going to be useful information. Maybe something like:

/* Allow for huge pages on HAVE_ARCH_HUGE_VMALLOC_FLAG arch's */