public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
To: linux-kernel@vger.kernel.org, linux-s390@vger.kernel.org
Cc: Heiko Carstens <heiko.carstens@de.ibm.com>,
	Martin Schwidefsky <schwidefsky@de.ibm.com>
Subject: [patch 14/34] convert/optimize csum_fold() to C
Date: Fri, 14 Aug 2009 13:25:31 +0200	[thread overview]
Message-ID: <20090814112615.846335152@de.ibm.com> (raw)
In-Reply-To: 20090814112517.982007860@de.ibm.com

[-- Attachment #1: 113-csum-fold.diff --]
[-- Type: text/plain, Size: 2058 bytes --]

From: Heiko Carstens <heiko.carstens@de.ibm.com>

In the meantime gcc generates better code than the old inline
assemblies do. Original inline assembly results in:

lr	%r1,%r2
sr	%r3,%r3
lr	%r2,%r1
srdl	%r2,16
alr	%r2,%r3
alr	%r1,%r2
srl	%r1,16
xilf	%r1,65535
llghr	%r2,%r1
br	%r14

Out of the C code gcc generates this:

rll	%r1,%r2,16
ar	%r1,%r2
srl	%r1,16
xilf	%r1,65535
llghr	%r2,%r1
br	%r14

In addition we don't have any static register allocations anymore and
gcc is free to shuffle instructions around for better pipeline usage.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---

 arch/s390/include/asm/checksum.h |   25 ++++---------------------
 1 file changed, 4 insertions(+), 21 deletions(-)

Index: quilt-2.6/arch/s390/include/asm/checksum.h
===================================================================
--- quilt-2.6.orig/arch/s390/include/asm/checksum.h
+++ quilt-2.6/arch/s390/include/asm/checksum.h
@@ -78,28 +78,11 @@ csum_partial_copy_nocheck (const void *s
  */
 static inline __sum16 csum_fold(__wsum sum)
 {
-#ifndef __s390x__
-	register_pair rp;
+	u32 csum = (__force u32) sum;
 
-	asm volatile(
-		"	slr	%N1,%N1\n"	/* %0 = H L */
-		"	lr	%1,%0\n"	/* %0 = H L, %1 = H L 0 0 */
-		"	srdl	%1,16\n"	/* %0 = H L, %1 = 0 H L 0 */
-		"	alr	%1,%N1\n"	/* %0 = H L, %1 = L H L 0 */
-		"	alr	%0,%1\n"	/* %0 = H+L+C L+H */
-		"	srl	%0,16\n"	/* %0 = H+L+C */
-		: "+&d" (sum), "=d" (rp) : : "cc");
-#else /* __s390x__ */
-	asm volatile(
-		"	sr	3,3\n"		/* %0 = H*65536 + L */
-		"	lr	2,%0\n"		/* %0 = H L, 2/3 = H L / 0 0 */
-		"	srdl	2,16\n"		/* %0 = H L, 2/3 = 0 H / L 0 */
-		"	alr	2,3\n"		/* %0 = H L, 2/3 = L H / L 0 */
-		"	alr	%0,2\n"		/* %0 = H+L+C L+H */
-		"	srl	%0,16\n"	/* %0 = H+L+C */
-		: "+&d" (sum) : : "cc", "2", "3");
-#endif /* __s390x__ */
-	return (__force __sum16) ~sum;
+	csum += (csum >> 16) + (csum << 16);
+	csum >>= 16;
+	return (__force __sum16) ~csum;
 }
 
 /*

-- 
blue skies,
   Martin.

"Reality continues to ruin my life." - Calvin.


  parent reply	other threads:[~2009-08-14 11:33 UTC|newest]

Thread overview: 35+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2009-08-14 11:25 [patch 00/34] s390 patch queue for the merge window of 2.6.32 Martin Schwidefsky
2009-08-14 11:25 ` [patch 01/34] cio: fix ineffective verify event Martin Schwidefsky
2009-08-14 11:25 ` [patch 02/34] cio: move scsw helper functions to header file Martin Schwidefsky
2009-08-14 11:25 ` [patch 03/34] cio: consolidate subchannel intparm reset Martin Schwidefsky
2009-08-14 11:25 ` [patch 04/34] cio: fix not oper handling after failed line processing Martin Schwidefsky
2009-08-14 11:25 ` [patch 05/34] cio: fix double free after failed device initialization Martin Schwidefsky
2009-08-14 11:25 ` [patch 06/34] qdio: continue polling if the queue is not finished Martin Schwidefsky
2009-08-14 11:25 ` [patch 07/34] cio: ensure to hold a reference for deferred deregistration Martin Schwidefsky
2009-08-14 11:25 ` [patch 08/34] qdio: remove limited number of debugfs entries Martin Schwidefsky
2009-08-14 11:25 ` [patch 09/34] dasd: fail requests when device state is less then ready Martin Schwidefsky
2009-08-14 11:25 ` [patch 10/34] dasd: optimize cpu usage in goodcase Martin Schwidefsky
2009-08-14 11:25 ` [patch 11/34] dasd: fix message naming Martin Schwidefsky
2009-08-14 11:25 ` [patch 12/34] drivers/s390: put NULL test before dereference Martin Schwidefsky
2009-08-14 11:25 ` [patch 13/34] introduce get_clock_monotonic Martin Schwidefsky
2009-08-14 11:25 ` Martin Schwidefsky [this message]
2009-08-14 11:25 ` [patch 15/34] improve mcount code Martin Schwidefsky
2009-08-14 11:25 ` [patch 16/34] atomic ops: add effecient atomic64 support for 31 bit Martin Schwidefsky
2009-08-14 11:25 ` [patch 17/34] atomic ops: small cleanups Martin Schwidefsky
2009-08-14 11:25 ` [patch 18/34] hibernation: remove dead file Martin Schwidefsky
2009-08-14 11:25 ` [patch 19/34] hibernation: merge files and move to kernel/ Martin Schwidefsky
2009-08-14 11:25 ` [patch 20/34] proper use of device register Martin Schwidefsky
2009-08-14 11:25 ` [patch 21/34] tape: use init_timer_on_stack() rather than init_timer() Martin Schwidefsky
2009-08-14 11:25 ` [patch 22/34] kernel: Append scpdata to kernel boot command line Martin Schwidefsky
2009-08-14 11:25 ` [patch 23/34] kernel: Convert upper case scpdata to lower case Martin Schwidefsky
2009-08-14 11:25 ` [patch 24/34] move (io|sysc)_restore_trace_psw into .data section Martin Schwidefsky
2009-08-14 11:25 ` [patch 25/34] Use macros for .data.page_aligned Martin Schwidefsky
2009-08-14 11:25 ` [patch 26/34] clean up linker script using new linker script macros Martin Schwidefsky
2009-08-14 11:25 ` [patch 27/34] kernel: always keep machine flags in lowcore Martin Schwidefsky
2009-08-14 11:25 ` [patch 28/34] remove unused irq_cpustat_t defintion Martin Schwidefsky
2009-08-14 11:25 ` [patch 29/34] add call home support Martin Schwidefsky
2009-08-14 11:25 ` [patch 30/34] hypfs: remove useless variable qname Martin Schwidefsky
2009-08-14 11:25 ` [patch 31/34] vmur: Invalid allocation sequence for vmur class Martin Schwidefsky
2009-08-14 11:25 ` [patch 32/34] xpram: Remove checksum validation for suspend/resume Martin Schwidefsky
2009-08-14 11:25 ` [patch 33/34] zcrypt: Use spin_lock_bh in suspend callback Martin Schwidefsky
2009-08-14 11:25 ` [patch 34/34] kernel: Set preferred s390 console based on conmode Martin Schwidefsky

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=20090814112615.846335152@de.ibm.com \
    --to=schwidefsky@de.ibm.com \
    --cc=heiko.carstens@de.ibm.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-s390@vger.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