xfs
[Top] [All Lists]

Re: [PATCH 3/4] mm: numa: Mark huge PTEs young when clearing NUMA hintin

To: Mel Gorman <mgorman@xxxxxxx>
Subject: Re: [PATCH 3/4] mm: numa: Mark huge PTEs young when clearing NUMA hinting faults
From: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx>
Date: Sat, 7 Mar 2015 10:42:29 -0800
Cc: Dave Chinner <david@xxxxxxxxxxxxx>, Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>, Ingo Molnar <mingo@xxxxxxxxxx>, Aneesh Kumar <aneesh.kumar@xxxxxxxxxxxxxxxxxx>, Linux Kernel Mailing List <linux-kernel@xxxxxxxxxxxxxxx>, Linux-MM <linux-mm@xxxxxxxxx>, xfs@xxxxxxxxxxx, ppc-dev <linuxppc-dev@xxxxxxxxxxxxxxxx>
Delivered-to: xfs@xxxxxxxxxxx
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=gmail.com; s=20120113; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=JosSfpBK6HOsRqK3sAR2tvnV6rh0MUEXcD3U9ozX5wc=; b=abGIq3sGfA9u2KUDuvc/zlEEV0C4RdKBmeCy+yXZpzPaYnGEKg5pbcqy0/K3XdHzQE L6p+1dvQeOUVEPqrqr4OuOwZNj5yf0X4pKtejvT9YIPvPNzJ+8oCqfHzXCAFZyKV2wJ/ 09wcEnig/OVy5NOxkrtCYkIzTyBAqLE3c5OP5OW8Wqt8sdgQsLRLa0qod2hBtv9AcptG DkFA6DYhkRJsHhShCpdhLZSJ7LPVMoV0J+UCEQi2GnSeliuNVgJfB+S12R4SJJlS1kG+ Rnj5ojLizxI9rwU7NfVzWsFaiwJKpoYZvCVdymN9bCyHbOhdAlL7CsSrMp5BWlkBIVO2 efag==
Dkim-signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=linux-foundation.org; s=google; h=mime-version:sender:in-reply-to:references:date:message-id:subject :from:to:cc:content-type; bh=JosSfpBK6HOsRqK3sAR2tvnV6rh0MUEXcD3U9ozX5wc=; b=SbzNKvo2AHFAPD+P4p9JndpUBdp89H/VZbHV3Fsqsw34pU3fbWatLDZtE3IqK1ZUE+ HYYpaRwTY8G3zz3t5MxXZH9NK33nm3DW7OEGiE3jqsVQPLsG0njq3yQyB4x0MjuywkIm udAexF/aKHKJ/gcGp8qMBdQkfmqCL+86R5Eio=
In-reply-to: <CA+55aFwSQgrYqfXPr6RPvQ+8OJfexXJRY_GVEKg5QtB2t38cWA@xxxxxxxxxxxxxx>
References: <1425741651-29152-1-git-send-email-mgorman@xxxxxxx> <1425741651-29152-4-git-send-email-mgorman@xxxxxxx> <CA+55aFwSQgrYqfXPr6RPvQ+8OJfexXJRY_GVEKg5QtB2t38cWA@xxxxxxxxxxxxxx>
Sender: linus971@xxxxxxxxx
On Sat, Mar 7, 2015 at 10:33 AM, Linus Torvalds
<torvalds@xxxxxxxxxxxxxxxxxxxx> wrote:
>
>             Completely untested, but that "just
> or in the new protection bits" is what pnf_pte() does just a few lines
> above this.

Hmm. Looking at this, we do *not* want to set _PAGE_ACCESSED when we
turn a page into PROT_NONE or mark it for numa faulting. Nor do we
want to set it for mprotect for random pages that we haven't actually
accessed, just changed the protections for.

So my patch was obviously wrong, and I should feel bad for suggesting
it. I'm a moron, and my expectations that "pte_modify()" would just
take the accessed bit from the vm_page_prot field was stupid and
wrong.

Mel's patch is the right thing to do.

                                Linus

<Prev in Thread] Current Thread [Next in Thread>