From: "Elliott, Robert (Persistent Memory)" <elliott@hpe.com>
To: Matt Fleming <matt@codeblueprint.co.uk>,
Andy Shevchenko <andy.shevchenko@gmail.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
Ingo Molnar <mingo@redhat.com>, "H. Peter Anvin" <hpa@zytor.com>,
"x86@kernel.org" <x86@kernel.org>,
"linux-efi@vger.kernel.org" <linux-efi@vger.kernel.org>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: RE: [PATCH 4/4] x86/efi: print size and base in binary units in efi_print_memmap
Date: Fri, 8 Jan 2016 16:38:17 +0000 [thread overview]
Message-ID: <94D0CD8314A33A4D9D801C0FE68B40295BF07E19@G4W3202.americas.hpqcorp.net> (raw)
In-Reply-To: <20160108121921.GI2532@codeblueprint.co.uk>
> -----Original Message-----
> From: Matt Fleming [mailto:matt@codeblueprint.co.uk]
> Sent: Friday, January 8, 2016 6:19 AM
> To: Andy Shevchenko <andy.shevchenko@gmail.com>
> Cc: Elliott, Robert (Persistent Memory) <elliott@hpe.com>; Thomas Gleixner
> <tglx@linutronix.de>; Ingo Molnar <mingo@redhat.com>; H. Peter Anvin
> <hpa@zytor.com>; x86@kernel.org; linux-efi@vger.kernel.org; linux-
> kernel@vger.kernel.org
> Subject: Re: [PATCH 4/4] x86/efi: print size and base in binary units in
> efi_print_memmap
>
> On Sun, 27 Dec, at 04:35:12PM, Andy Shevchenko wrote:
> > On Mon, Dec 21, 2015 at 6:16 PM, Matt Fleming <matt@codeblueprint.co.uk>
> wrote:
> > >> diff --git a/arch/x86/platform/efi/efi.c
> b/arch/x86/platform/efi/efi.c
> > >> index 635a955..030ba91 100644
> > >> --- a/arch/x86/platform/efi/efi.c
> > >> +++ b/arch/x86/platform/efi/efi.c
> > >> @@ -222,6 +222,25 @@ int __init efi_memblock_x86_reserve_range(void)
> > >> return 0;
> > >> }
> > >>
> > >> +char * __init efi_size_format(char *buf, size_t size, u64 bytes)
> > >> +{
> > >> + if (!bytes || (bytes & 0x3ff))
> > >> + snprintf(buf, size, "%llu B", bytes);
> > >> + else if (bytes & 0xfffff)
> > >> + snprintf(buf, size, "%llu KiB", bytes >> 10);
> > >> + else if (bytes & 0x3fffffff)
> > >> + snprintf(buf, size, "%llu MiB", bytes >> 20);
> > >> + else if (bytes & 0xffffffffff)
> > >> + snprintf(buf, size, "%llu GiB", bytes >> 30);
> > >> + else if (bytes & 0x3ffffffffffff)
> > >> + snprintf(buf, size, "%llu TiB", bytes >> 40);
> > >> + else if (bytes & 0xfffffffffffffff)
> > >> + snprintf(buf, size, "%llu PiB", bytes >> 50);
> > >> + else
> > >> + snprintf(buf, size, "%llu EiB", bytes >> 60);
> > >> + return buf;
> >
> > For me it looks like ffs with name in the table can be used.
>
> Could you provide a patch?
I think this is functionally equivalent:
#include <string.h>
char * efi_size_format_ffsl(char *buf, size_t size, u64 bytes)
{
if (!bytes || ffsl(bytes) < 10)
snprintf(buf, size, "%llu B", bytes);
else if (ffsl(bytes) < 20)
snprintf(buf, size, "%llu KiB", bytes >> 10);
else if (ffsl(bytes) < 30)
snprintf(buf, size, "%llu MiB", bytes >> 20);
else if (ffsl(bytes) < 40)
snprintf(buf, size, "%llu GiB", bytes >> 30);
else if (ffsl(bytes) < 50)
snprintf(buf, size, "%llu TiB", bytes >> 40);
else if (ffsl(bytes) < 60)
snprintf(buf, size, "%llu PiB", bytes >> 50);
else
snprintf(buf, size, "%llu EiB", bytes >> 60);
return buf;
}
Compiled as a user program with gcc -O2, the original results
in mov and testq instructions:
movq %rdi, %rbx
je .L2
testl $1023, %edx
jne .L2
testl $1048575, %edx
jne .L15
testl $1073741823, %edx
jne .L16
movabsq $1099511627775, %rax
testq %rax, %rdx
jne .L17
movabsq $1125899906842623, %rax
testq %rax, %rdx
jne .L18
movabsq $1152921504606846975, %rax
movq %rdx, %rcx
testq %rax, %rdx
jne .L19
while the ffs version uses bit scan forward (bsfq)
and only needs cmpl instructions since the values
are smaller:
movq %rdi, %rbx
je .L21
bsfq %rdx, %rcx
addq $1, %rcx
cmpl $9, %ecx
jle .L21
cmpl $19, %ecx
jle .L33
cmpl $29, %ecx
jle .L34
cmpl $39, %ecx
.p2align 4,,2
jle .L35
cmpl $49, %ecx
.p2align 4,,2
jle .L36
cmpl $59, %ecx
.p2align 4,,2
jle .L37
The kernel offers ffs(int x) but not ffsl(), and it
uses inline assembly for one of these:
bsfl
bsfl, cmovzl
bsfl, jnz, movl
I don't know which code is the most efficient.
---
Robert Elliott, HPE Persistent Memory
next prev parent reply other threads:[~2016-01-08 16:38 UTC|newest]
Thread overview: 25+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-12-18 1:28 Robert Elliott
2015-12-18 1:28 ` [PATCH 1/4] x86/efi: show actual ending addresses in efi_print_memmap Robert Elliott
2015-12-21 15:50 ` Matt Fleming
2015-12-21 16:06 ` Matt Fleming
2015-12-22 20:08 ` Elliott, Robert (Persistent Memory)
2015-12-23 12:44 ` Matt Fleming
2015-12-21 16:44 ` Elliott, Robert (Persistent Memory)
2015-12-23 12:47 ` Matt Fleming
2015-12-24 1:07 ` [PATCH v2 " Robert Elliott
2016-01-08 12:04 ` Matt Fleming
2015-12-18 1:28 ` [PATCH 2/4] efi: add NV memory attribute Robert Elliott
2015-12-21 15:54 ` Matt Fleming
2015-12-18 1:28 ` [PATCH 3/4] efi: add Persistent Memory type name Robert Elliott
2016-01-08 12:20 ` Matt Fleming
2015-12-18 1:28 ` [PATCH 4/4] x86/efi: print size and base in binary units in efi_print_memmap Robert Elliott
2015-12-21 16:16 ` Matt Fleming
2015-12-23 0:11 ` Elliott, Robert (Persistent Memory)
2015-12-23 15:52 ` Matt Fleming
2015-12-27 14:35 ` Andy Shevchenko
2016-01-08 12:19 ` Matt Fleming
2016-01-08 16:38 ` Elliott, Robert (Persistent Memory) [this message]
2016-01-08 16:44 ` Andy Shevchenko
2016-01-08 16:39 ` Andy Shevchenko
2016-01-11 14:09 ` Matt Fleming
2016-01-12 13:17 ` Andy Shevchenko
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=94D0CD8314A33A4D9D801C0FE68B40295BF07E19@G4W3202.americas.hpqcorp.net \
--to=elliott@hpe.com \
--cc=andy.shevchenko@gmail.com \
--cc=hpa@zytor.com \
--cc=linux-efi@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=matt@codeblueprint.co.uk \
--cc=mingo@redhat.com \
--cc=tglx@linutronix.de \
--cc=x86@kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).