[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <149c73af-f3a5-2704-ba73-5e2fbbb4749e@oracle.com>
Date: Mon, 24 Oct 2016 11:31:14 -0700
From: tndave <tushar.n.dave@...cle.com>
To: David Miller <davem@...emloft.net>
Cc: chris.hyser@...cle.com, sowmini.varadhan@...cle.com,
vgupta@...opsys.com, robin.murphy@....com, jroedel@...e.de,
egtvedt@...fundet.no, krzk@...nel.org, rkuo@...eaurora.org,
sparclinux@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/6] sparc64: Add FORCE_MAX_ZONEORDER and default to 13
On 10/24/2016 10:45 AM, David Miller wrote:
> From: Tushar Dave <tushar.n.dave@...cle.com>
> Date: Mon, 10 Oct 2016 11:12:02 -0700
>
>> From: Dave Kleikamp <dave.kleikamp@...cle.com>
>>
>> This change allows ATU (new IOMMU) in SPARC systems to request
>> large (32M) contiguous memory during boot for creating IOTSB backing
>> store.
>>
>> Signed-off-by: Dave Kleikamp <dave.kleikamp@...cle.com>
>> Signed-off-by: Tushar Dave <tushar.n.dave@...cle.com>
>
> If you need 32MB allocations, wouldn't a max zone order of 12 be
> sufficient? (8K << 12 == 32MB)
Yes, but config FORCE_MAX_ZONEORDER is actually maximum order plus one.
For example, a value of 13 means that the largest free memory block is
2^12 pages. For 32MB we need 2^12 8K pages, hence FORCE_MAX_ZONEORDER
must be MAX_ORDER + 1 i.e. 13.
Thanks.
-Tushar
>
Powered by blists - more mailing lists