arm: Adding support for atomic half word exchange
PATCH v3 1/1 arm: Adding support for atomic half word exchange
Hello Will,
Thank you so much for your patience. My sincere apologies with the issues with formatting. I have taken care of them and hope now things are ok.
- Included patchesATarm.linux.org.uk.
- Removed the changelog from commit message.
- Moved the patch to "superseded" in patch system.
- Corrected the truncated Signed-off line.
KernelVersion: 4.3-rc4
Since support for half-word atomic exchange was not there and Qspinlock on ARM requires it, modified __xchg() to add support for that as well. ARMv6 and lower does not support ldrex{b,h} so, added a guard code to prevent build breaks.
--- arch/arm/include/asm/cmpxchg.h | 12 ++++++++++++ 1 file changed, 12 insertions(+)
arch/arm/include/asm/cmpxchg.h +12 -0
@@ -39,6 +39,7 @@ static inline unsigned long __xchg(unsigned long x, volatile void *ptr, int sizeswitch (size) {#if __LINUX_ARM_ARCH__ >=6 +#ifndef CONFIG_CPU_V6 /* MIN ARCH >= V6K */ case 1 : asmvolatile ("@ __xchg1\n" "1: ldrexb %0, [%3]\n"
@@ -49,6 +50,17 @@ static inline unsigned long __xchg(unsigned long x, volatile void *ptr, int size : "r" (x), "r" (ptr) : "memory", "cc");break ; +case 2 : + asmvolatile ("@ __xchg2\n" +"1: ldrexh %0, [%3]\n" +" strexh %1, %2, [%3]\n" +" teq %1, #0\n" +" bne 1b" + :"=&r" (ret),"=&r" (tmp) + :"r" (x),"r" (ptr) + :"memory" ,"cc" ); +break ; +#endif case 4 : asmvolatile ("@ __xchg4\n" "1: ldrex %0, [%3]\n"
Signature · 8 lines
1.9.1 Sarbojit ------- Original Message ------- Sender : Will Deacon<[email protected]> Date : Oct 08, 2015 22:26 (GMT+05:30) Title : Re: Re: Re: Re: Re: [PATCH v3] arm: Adding support for atomic half word exchange
Quoted 3 lines · Sarbojit Ganguly
Please have a look at this patch, please let me know if any modification is required.
I have also submitted the same in your patch system.
There are some problems with the version in the patch system[1]:
(1) You still have the changelog in the commit message (I asked you to remove this last time).
(2) The commit message isn't line-wrapper appropriately
(3) Your Signed-off-by line is truncated
The patch system does actually support git, so you can add:
KernelVersion: 4.3-rc4
or whatever kernel you based your patch on somewhere after your SoB but before the diff and then mail it to patchesATarm.linux.org.uk. You probably also want to move the current patch to "superseded" so Russell doesn't end up with two copies.
Will
[1] http://www.arm.linux.org.uk/developer/patches/viewpatch.php?id=8442/1
Quoted 10 lines · Sarbojit Ganguly
Hello Will,
Thank you so much for your patience. My sincere apologies with the issues with formatting. I have taken care of them and hope now things are ok.
- Included patchesATarm.linux.org.uk.
- Removed the changelog from commit message.
- Moved the patch to "superseded" in patch system.
- Corrected the truncated Signed-off line.
Quoted 8 lines
KernelVersion: 4.3-rc4
Since support for half-word atomic exchange was not there and Qspinlock on ARM requires it, modified
__xchg()to add support for that as well. ARMv6 and lower does not support ldrex{b,h} so, added a guard code to prevent build breaks.Signed-off-by Sarbojit Ganguly <[email protected]>---
The commit message is everything above the "---" line. It starts with the subject line, a blank line, and then "Hello Will,".
I'll fix up the commit message this time, but please take note for future patches.
You're also the first one to send a patch where the body _inside_ the base64 container has CRLF line endings...
Quoted 69 lines
arch/arm/include/asm/cmpxchg.h | 12 ++++++++++++ 1 file changed, 12 insertions(+)arch/arm/include/asm/cmpxchg.h +12
-0@@ -39,6 +39,7 @@ static inline unsigned long __xchg(unsigned long x, volatile void *ptr, int sizeswitch (size) {#if __LINUX_ARM_ARCH__ >=6 +#ifndef CONFIG_CPU_V6 /* MIN ARCH >= V6K */ case 1 : asmvolatile ("@ __xchg1\n" "1: ldrexb %0, [%3]\n"@@ -49,6 +50,17 @@ static inline unsigned long __xchg(unsigned long x, volatile void *ptr, int size : "r" (x), "r" (ptr) : "memory", "cc");break ; +case 2 : + asmvolatile ("@ __xchg2\n" +"1: ldrexh %0, [%3]\n" +" strexh %1, %2, [%3]\n" +" teq %1, #0\n" +" bne 1b" + :"=&r" (ret),"=&r" (tmp) + :"r" (x),"r" (ptr) + :"memory" ,"cc" ); +break ; +#endif case 4 : asmvolatile ("@ __xchg4\n" "1: ldrex %0, [%3]\n"Signature · 8 lines
1.9.1 Sarbojit ------- Original Message ------- Sender : Will Deacon<[email protected]> Date : Oct 08, 2015 22:26 (GMT+05:30) Title : Re: Re: Re: Re: Re: [PATCH v3] arm: Adding support for atomic half word exchangeQuoted 3 lines · Sarbojit Ganguly
Please have a look at this patch, please let me know if any modification is required.
I have also submitted the same in your patch system.There are some problems with the version in the patch system[1]:
(1) You still have the changelog in the commit message (I asked you to remove this last time).
(2) The commit message isn't line-wrapper appropriately
(3) Your Signed-off-by line is truncated
The patch system does actually support git, so you can add:
KernelVersion: 4.3-rc4
or whatever kernel you based your patch on somewhere after your SoB but before the diff and then mail it to patchesATarm.linux.org.uk. You probably also want to move the current patch to "superseded" so Russell doesn't end up with two copies.
Will
[1] http://www.arm.linux.org.uk/developer/patches/viewpatch.php?id=8442/1
Signature · 2 lines
FTTC broadband for 0.8mile line: currently at 9.6Mbps down 400kbps up according to speedtest.net.