Re: [tip:x86/asm] x86: Pack function addresses tightly as well
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
- Subject: Re: [tip:x86/asm] x86: Pack function addresses tightly as well
- From: Ingo Molnar <mingo@xxxxxxxxxx>
- Date: Sun, 17 May 2015 09:30:08 +0200
- Cc: Andy Lutomirski <luto@xxxxxxxxxxxxxx>, Davidlohr Bueso <dave@xxxxxxxxxxxx>, Peter Anvin <hpa@xxxxxxxxx>, Denys Vlasenko <dvlasenk@xxxxxxxxxx>, Linux Kernel Mailing List <linux-kernel@xxxxxxxxxxxxxxx>, Tim Chen <tim.c.chen@xxxxxxxxxxxxxxx>, Borislav Petkov <bp@xxxxxxxxx>, Peter Zijlstra <peterz@xxxxxxxxxxxxx>, "Chandramouleeswaran, Aswin" <aswin@xxxxxx>, Peter Zijlstra <a.p.zijlstra@xxxxxxxxx>, Brian Gerst <brgerst@xxxxxxxxx>, Paul McKenney <paulmck@xxxxxxxxxxxxxxxxxx>, Thomas Gleixner <tglx@xxxxxxxxxxxxx>, Jason Low <jason.low2@xxxxxx>, "linux-tip-commits@xxxxxxxxxxxxxxx" <linux-tip-commits@xxxxxxxxxxxxxxx>
- In-reply-to: <20150517070929.GA23344@gmail.com>
- User-agent: Mutt/1.5.23 (2014-03-12)
* Ingo Molnar <mingo@xxxxxxxxxx> wrote:
> The median function size is around 1 cacheline (64-byte one), ~80%
> fitting into two cachelines, with a big peak for very small
> functions that make up something like 20% of all functions [...]
Correction:
32% of kernel functions fit into a single cacheline,
55% fit into two cachelines,
70% into three cachelines,
76% into four cachelines
so the tail is longer than my quick read of the graph suggested.
OTOH, probability of use is biased towards smaller functions: we tend
to use smaller, facility functions more frequently.
Thanks,
Ingo
--
To unsubscribe from this list: send the line "unsubscribe linux-tip-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
[Index of Archives]
[Linux Stable Commits]
[Linux Stable Kernel]
[Linux Kernel]
[Linux USB Devel]
[Linux Video &Media]
[Linux Audio Users]
[Yosemite News]
[Linux SCSI]