arm: prevent using movt/movw address loads
PATCH 1/1 arm: prevent using movt/movw address loads
The movt/movw instruction can be used to hardcode an memory location in the instruction itself. The linker starts complaining about this if the compiler decides to do so: "relocation R_ARM_MOVW_ABS_NC against `a local symbol' can not be used" and it is not support by U-boot as well. Prevent their use by requiring word relocations. This allows u-boot to be build at other optimalization levels then -Os.
--- arch/arm/config.mk | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
arch/arm/config.mk +6 -2
@@ -94,7 +94,11 @@ PLATFORM_RELFLAGS += -fno-optimize-sibling-calls endif endif -# check that only R_ARM_RELATIVE relocations are generated ifneq ($(CONFIG_SPL_BUILD),y) -ALL-y += checkarmreloc +# Check that only R_ARM_RELATIVE relocations are generated. +ALL-y += checkarmreloc +# The movt / movw can hardcode 16 bit parts of the addresses in the +# instruction. Relocation is not supported for that case, so disable +# such usage by requiring word relocations. +PLATFORM_CPPFLAGS += $(call cc-option, -mword-relocations) endif
Signature · 1 line
1.8.1.2
PATCH 1/1 arm: prevent using movt/movw address loads
Hi, experts:
Quoted 10 lines
-# check that only
R_ARM_RELATIVErelocations are generated ifneq ($(CONFIG_SPL_BUILD),y)arch/arm/config.mk +6
-1-ALL-y += checkarmreloc +# Check that only R_ARM_RELATIVE relocations are generated. +ALL-y += checkarmreloc +# The movt / movw can hardcode 16 bit parts of the addresses in the +# instruction. Relocation is not supported for that case, so disable +# such usage by requiring word relocations. +PLATFORM_CPPFLAGS += $(call cc-option, -mword-relocations)endif
Jeroen's patch is very simple.
So, is there any side-effect?
If not, why not add it into 2013.10 release version? :)
Best wishes,
PATCH 1/1 arm: prevent using movt/movw address loads
Quoted 3 lines · TigerLiu at viatech.com.cn
Jeroen's patch is very simple.
So, is there any side-effect?
Not that I am aware of.
Quoted 1 line
If not, why not add it into 2013.10 release version? :)
That is up to Albert and Tom.
Regards,
Jeroen
PATCH 1/1 arm: prevent using movt/movw address loads
In reply to Jeroen HofsteeQuoted 9 lines · Jeroen Hofstee
The movt/movw instruction can be used to hardcode an memory location in the instruction itself. The linker starts complaining about this if the compiler decides to do so: "relocation
R_ARM_MOVW_ABS_NCagainst `a local symbol' can not be used" and it is not support by U-boot as well. Prevent their use by requiring word relocations. This allows u-boot to be build at other optimalization levels then -Os.
Quoted 6 lines
Signed-off-by Jeroen Hofstee <[email protected]>Cc TigerLiu at viatech.com.cnCc Albert ARIBAUD <[email protected]>--- arch/arm/config.mk | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)
Is this also something we need for llvm? I am hesitant here because as Wolfgang points out, -O0 is usually the wrong way to debug a problem and I'll add we're well into the age where debuggers work just fine with optimized code. If there's some -O2 enabled gcc flag we want because of a measurable performance win, we should add it specifically to -Os.
Signature · 8 lines
Tom -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 836 bytes Desc: Digital signature URL: <http://lists.denx.de/pipermail/u-boot/attachments/20130919/d8945885/attachment.pgp>
PATCH 1/1 arm: prevent using movt/movw address loads
Hello Tom,
Quoted 11 lines · Tom Rini
Quoted 9 lines · Jeroen Hofstee
The movt/movw instruction can be used to hardcode an memory location in the instruction itself. The linker starts complaining about this if the compiler decides to do so: "relocation
R_ARM_MOVW_ABS_NCagainst `a local symbol' can not be used" and it is not support by U-boot as well. Prevent their use by requiring word relocations. This allows u-boot to be build at other optimalization levels then -Os.
Quoted 7 lines
Quoted 6 lines
Signed-off-by Jeroen Hofstee <[email protected]>Cc TigerLiu at viatech.com.cnCc Albert ARIBAUD <[email protected]>--- arch/arm/config.mk | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)Is this also something we need for llvm?
You guessed that right, for clang actually (llvm has already been taught to not emit movw/movt pairs, when requested not to do so). So with the -mword-relocations || present I can teach clang to tell llvm not to do it.
I am not aware of any reason why gcc could not decide to do the same in future releases. A pointer comparison e.g. is of exactly the same size (afaik). In this case U-boot will no longer compile without mentioned flag.
Quoted 4 lines
I am hesitant here because as
Wolfgang points out, -O0 is usually the wrong way to debug a problem and I'll add we're well into the age where debuggers work just fine with optimized code.
mmm, I don't share your concern here. Not that I disagree with what Wolfgang said, but since it is unrelated to the patch itself. What I read was that Wolfgang tried to explain to a ML poster without a proper name that it might be even harder at times to find a bug at -O0, since it is a different binary and that it is not considered a bug. I assume the fast majority of U-boot developers know these to debug things..
If you really have that little trust in U-boot developers a more proper way would be to actually create a make rule checking cflags and point them to a nice debugging document. And I really hope you don't do that ;)
One thing I can think of in favour of -O0 is for educational purposes. You can run u-boot in qemu then without the, at times weird optimized jumps, to get an idea about basic code flow.
Quoted 3 lines
If there's some -O2 enabled gcc flag we want because of a measurable performance win, we should add it specifically to -Os.
First of all the default -Os is unchanged and I have no intention to change it. -O2 won't build without the patch last time I checked ;)
Anyway, I like the flag since it helps to not special case clang and it guarantees builds with gcc at all optimisation levels, now and in the future. I don't care if it goes in this release or the next one.
Regards,
Jeroen
PATCH 1/1 arm: prevent using movt/movw address loads
Quoted 28 lines · Jeroen Hofstee
Hello Tom,
Quoted 11 lines · Tom Rini
Quoted 9 lines · Jeroen Hofstee
The movt/movw instruction can be used to hardcode an memory location in the instruction itself. The linker starts complaining about this if the compiler decides to do so: "relocation
R_ARM_MOVW_ABS_NCagainst `a local symbol' can not be used" and it is not support by U-boot as well. Prevent their use by requiring word relocations. This allows u-boot to be build at other optimalization levels then -Os.Quoted 7 lines
Quoted 6 lines
Signed-off-by Jeroen Hofstee <[email protected]>Cc TigerLiu at viatech.com.cnCc Albert ARIBAUD <[email protected]>--- arch/arm/config.mk | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)Is this also something we need for llvm?
You guessed that right, for clang actually (llvm has already been taught to not emit movw/movt pairs, when requested not to do so). So with the -mword-relocations || present I can teach clang to tell llvm not to do it.
Quoted 5 lines
I am not aware of any reason why gcc could not decide to do the same in future releases. A pointer comparison e.g. is of exactly the same size (afaik). In this case U-boot will no longer compile without mentioned flag.
OK.
[snip]
Quoted 11 lines
Quoted 3 lines
If there's some -O2 enabled gcc flag we want because of a measurable performance win, we should add it specifically to -Os.
First of all the default -Os is unchanged and I have no intention to change it. -O2 won't build without the patch last time I checked ;)
Anyway, I like the flag since it helps to not special case clang and it guarantees builds with gcc at all optimisation levels, now and in the future. I don't care if it goes in this release or the next one.
Right, I'm OK picking this patch up then, on the grounds of making clang/llvm work now, and potentially keeping a future gcc happy.
Signature · 8 lines
Tom -------------- next part -------------- A non-text attachment was scrubbed... Name: not available Type: application/pgp-signature Size: 836 bytes Desc: Digital signature URL: <http://lists.denx.de/pipermail/u-boot/attachments/20130920/3f1d0eca/attachment.pgp>
PATCH 1/1 arm: prevent using movt/movw address loads
In reply to Jeroen HofsteeQuoted 9 lines · Jeroen Hofstee
The movt/movw instruction can be used to hardcode an memory location in the instruction itself. The linker starts complaining about this if the compiler decides to do so: "relocation
R_ARM_MOVW_ABS_NCagainst `a local symbol' can not be used" and it is not support by U-boot as well. Prevent their use by requiring word relocations. This allows u-boot to be build at other optimalization levels then -Os.
Quoted 4 lines
Signed-off-by Jeroen Hofstee <[email protected]>Cc TigerLiu at viatech.com.cnCc Albert ARIBAUD <[email protected]>
This is useful I think. I'm not sure that -O0 works very well anymore (at least I need to make a few tweaks to use it), but -O1 is useful in some cases to provide better debugging.
PATCH 1/1 arm: prevent using movt/movw address loads
In reply to Jeroen HofsteeHi Jeroen,
Quoted 24 lines · Jeroen Hofstee
The movt/movw instruction can be used to hardcode an memory location in the instruction itself. The linker starts complaining about this if the compiler decides to do so: "relocation
R_ARM_MOVW_ABS_NCagainst `a local symbol' can not be used" and it is not support by U-boot as well. Prevent their use by requiring word relocations. This allows u-boot to be build at other optimalization levels then -Os.Signed-off-by Jeroen Hofstee <[email protected]>Cc TigerLiu at viatech.com.cnCc Albert ARIBAUD <[email protected]>--- arch/arm/config.mk | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)arch/arm/config.mk +0
-0@@ -94,7 +94,11 @@ PLATFORM_RELFLAGS += -fno-optimize-sibling-calls endif endif
arch/arm/config.mk PLATFORM_RELFLAGS += -fno-optimize-sibling-calls · line 104
Quoted 10 lines
arch/arm/config.mk +6
-2-# check that only R_ARM_RELATIVE relocations are generated ifneq ($(CONFIG_SPL_BUILD),y) -ALL-y += checkarmreloc +# Check that only R_ARM_RELATIVE relocations are generated. +ALL-y += checkarmreloc +# The movt / movw can hardcode 16 bit parts of the addresses in the +# instruction. Relocation is not supported for that case, so disable +# such usage by requiring word relocations. +PLATFORM_CPPFLAGS += $(call cc-option, -mword-relocations) endif
Applied as a bugfix to u-boot-arm/master, thanks!
Amicalement,
Signature · 1 line
Albert.