Pull request for branch gcc-remove-options

This is a very advanced option, and it seems, according to
http://choices.cs.uiuc.edu/exceptions.pdf that SJLJ exceptions aren't really interesting.

Users really interested by this can always use the BR2_EXTRA_GCC_CONFIG_OPTIONS is they want.

Signed-off-by Thomas Petazzoni <[email protected]>

---
 configs/arm_toolchain_defconfig      |    1 -
 configs/at91rm9200df_defconfig       |    1 -
 configs/at91sam9260dfc_defconfig     |    1 -
 configs/at91sam9260pf_defconfig      |    1 -
 configs/at91sam9261ek_defconfig      |    1 -
 configs/at91sam9263ek_defconfig      |    1 -
 configs/at91sam9g20dfc_defconfig     |    1 -
 configs/atngw100-base_defconfig      |    1 -
 configs/atngw100_defconfig           |    1 -
 configs/atstk1005_defconfig          |    1 -
 configs/atstk100x_defconfig          |    1 -
 configs/i386_defconfig               |    1 -
 configs/i686_defconfig               |    1 -
 configs/integrator926_defconfig      |    1 -
 configs/integrator926_huge_defconfig |    1 -
 configs/kb9202_defconfig             |    1 -
 configs/v100sc2_defconfig            |    1 -
 toolchain/gcc/Config.in              |    8 --------
 toolchain/gcc/Makefile.in            |    4 ----
 toolchain/gcc/gcc-uclibc-4.x.mk      |    2 --
 20 files changed, 0 insertions(+), 31 deletions(-)

configs/arm_toolchain_defconfig +0 -1

index 3062541..2ac7de3 100644
@@ -245,7 +245,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_FORTRAN is not set
 # BR2_INSTALL_LIBGCJ is not set

configs/at91rm9200df_defconfig +0 -1

index 98158b9..2a193cf 100644
@@ -250,7 +250,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_FORTRAN is not set
 # BR2_INSTALL_LIBGCJ is not set

configs/at91sam9260dfc_defconfig +0 -1

index f1d63df..98d3401 100644
@@ -267,7 +267,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_CROSS_CXX=y
 # BR2_GCC_CROSS_FORTRAN is not set

configs/at91sam9260pf_defconfig +0 -1

index 651041f..ec4005d 100644
@@ -173,7 +173,6 @@ BR2_GCC_VERSION_4_2_1=y
 # BR2_GCC_IS_SNAP is not set
 BR2_GCC_VERSION="4.2.1"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_CROSS_CXX=y
 # BR2_GCC_CROSS_FORTRAN is not set

configs/at91sam9261ek_defconfig +0 -1

index 0d2fa0b..7524c68 100644
@@ -270,7 +270,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_CROSS_CXX=y
 # BR2_GCC_CROSS_FORTRAN is not set

configs/at91sam9263ek_defconfig +0 -1

index 6b4f592..85ccede 100644
@@ -264,7 +264,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_FORTRAN is not set
 # BR2_INSTALL_LIBGCJ is not set

configs/at91sam9g20dfc_defconfig +0 -1

index 92171d5..57d72ae 100644
@@ -264,7 +264,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_FORTRAN is not set
 # BR2_INSTALL_LIBGCJ is not set

configs/atngw100-base_defconfig +0 -1

index f5eba61..270c056 100644
@@ -196,7 +196,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 # BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE is not set
 BR2_GCC_VERSION="4.2.2-avr32-2.1.5"
 # BR2_TOOLCHAIN_SYSROOT is not set
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_SHARED_LIBGCC is not set
 

configs/atngw100_defconfig +0 -1

index d1db2da..a78372d 100644
@@ -174,7 +174,6 @@ BR2_GCC_VERSION_4_2_2_AVR32_2_1_5=y
 BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_VERSION="4.2.2-avr32-2.1.5"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_SHARED_LIBGCC=y
 BR2_GCC_ENABLE_TLS=y

configs/atstk1005_defconfig +0 -1

index b90761b..451e248 100644
@@ -200,7 +200,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 # BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE is not set
 BR2_GCC_VERSION="4.2.2-avr32-2.1.5"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_SHARED_LIBGCC=y
 

configs/atstk100x_defconfig +0 -1

index c12ccf8..e39e608 100644
@@ -195,7 +195,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 # BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE is not set
 BR2_GCC_VERSION="4.2.2-avr32-2.1.5"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_SHARED_LIBGCC=y
 

configs/i386_defconfig +0 -1

index 65825be..026e206 100644
@@ -195,7 +195,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.4.0"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_FORTRAN is not set
 # BR2_INSTALL_LIBGCJ is not set

configs/i686_defconfig +0 -1

index 31ce604..884ec7d 100644
@@ -195,7 +195,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.4.0"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_FORTRAN is not set
 # BR2_INSTALL_LIBGCJ is not set

configs/integrator926_defconfig +0 -1

index 2eb22c5..4ec53f9 100644
@@ -199,7 +199,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 # BR2_GCC_CROSS_CXX is not set
 # BR2_GCC_CROSS_FORTRAN is not set

configs/integrator926_huge_defconfig +0 -1

index e44ac5f..b5c34df 100644
@@ -199,7 +199,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE=y
 BR2_GCC_VERSION="4.3.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_CROSS_CXX=y
 # BR2_GCC_CROSS_FORTRAN is not set

configs/kb9202_defconfig +0 -1

index 40c6d18..1d0aa61 100644
@@ -201,7 +201,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 # BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE is not set
 BR2_GCC_VERSION="4.2.4"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_CROSS_CXX=y
 # BR2_GCC_CROSS_FORTRAN is not set

configs/v100sc2_defconfig +0 -1

index ca261ee..5ba2899 100644
@@ -170,7 +170,6 @@ BR2_GCC_SUPPORTS_SYSROOT=y
 # BR2_GCC_SUPPORTS_FINEGRAINEDMTUNE is not set
 BR2_GCC_VERSION="4.2.2"
 BR2_TOOLCHAIN_SYSROOT=y
-# BR2_GCC_USE_SJLJ_EXCEPTIONS is not set
 BR2_EXTRA_GCC_CONFIG_OPTIONS=""
 BR2_GCC_CROSS_CXX=y
 BR2_INSTALL_LIBSTDCPP=y

toolchain/gcc/Config.in +0 -8

index 31eba6b..cbdc026 100644
@@ -68,14 +68,6 @@ config BR2_GCC_VERSION
 	default "4.4.4"     if BR2_GCC_VERSION_4_4_X
 	default $BR2_GCC_SNAP_DATE if BR2_GCC_VERSION_SNAP
 
-config BR2_GCC_USE_SJLJ_EXCEPTIONS
-	bool "Enable setjmp/longjmp exceptions?"
-	help
-	  For some platforms, proper stack unwinding works perfectly,
-	  while other platforms must use setjmp/longjmp exceptions for
-	  proper stack unwinding during exception handling.  Most people
-	  can leave this set to n.
-
 config BR2_EXTRA_GCC_CONFIG_OPTIONS
 	string "Additional gcc options"
 	default ""

toolchain/gcc/Makefile.in +0 -4

index 9537e9d..b6ebca9 100644
@@ -13,10 +13,6 @@ else
 GCC_SNAP_DATE:=
 endif
 
-ifeq ($(BR2_GCC_USE_SJLJ_EXCEPTIONS),y)
-GCC_USE_SJLJ_EXCEPTIONS:=--enable-sjlj-exceptions
-endif
-
 ifeq ($(BR2_SOFT_FLOAT),y)
 SOFT_FLOAT_CONFIG_OPTION:=--with-float=soft
 ifeq ($(BR2_arm)$(BR2_armeb),y) # only set float-abi for arm

toolchain/gcc/gcc-uclibc-4.x.mk +0 -2

index dfc1790..d9269bd 100644
@@ -290,7 +290,6 @@ $(GCC_BUILD_DIR2)/.configured: $(GCC_SRC_DIR)/.patched $(GCC_STAGING_PREREQ)
 		$(GCC_DECIMAL_FLOAT) \
 		$(SOFT_FLOAT_CONFIG_OPTION) \
 		$(GCC_WITH_ABI) $(GCC_WITH_ARCH) $(GCC_WITH_TUNE) \
-		$(GCC_USE_SJLJ_EXCEPTIONS) \
 		$(DISABLE_LARGEFILE) \
 		$(EXTRA_GCC_CONFIG_OPTIONS) \
 		$(EXTRA_GCC2_CONFIG_OPTIONS) \

@@ -424,7 +423,6 @@ $(GCC_BUILD_DIR3)/.configured: $(GCC_BUILD_DIR3)/.prepared
 		$(GCC_DECIMAL_FLOAT) \
 		$(SOFT_FLOAT_CONFIG_OPTION) \
 		$(GCC_WITH_ABI) $(GCC_WITH_ARCH) $(GCC_WITH_TUNE) \
-		$(GCC_USE_SJLJ_EXCEPTIONS) \
 		$(DISABLE_LARGEFILE) \
 		$(EXTRA_GCC_CONFIG_OPTIONS) \
 		$(EXTRA_TARGET_GCC_CONFIG_OPTIONS) \

And of course you confirmed that the Lua package's error handling still works properly, complete with
same backtrace that is produced when SJLJ exceptions are enabled.

You did make that test didn't you?

Hello,

Quoted 3 lines · On Thu, 29 Jul 2010 05:38:06 -0500 "Michael S. Zick" <[email protected]>…

And of course you confirmed that the Lua package's error handling still works properly, complete with
same backtrace that is produced when SJLJ exceptions are enabled.

No, I did not.

Quoted 1 line

You did make that test didn't you?

Why do you need to be so aggressive ? It is possible that removing this option was a mistake, but there's *really* no need to be so aggressive. It only reduces the motivation to look at the problem...

So, trying to be constructive, how can I test that Lua still produces correct backtraces ? (I'm not a Lua programmer at all).

os.exit("String Value")

Thanks,

Thomas

Quoted 9 lines · Thomas Petazzoni

Hello,

Quoted 3 lines · On Thu, 29 Jul 2010 05:38:06 -0500 "Michael S. Zick" <[email protected]>…

And of course you confirmed that the Lua package's error handling still works properly, complete with
same backtrace that is produced when SJLJ exceptions are enabled.

Quoted 8 lines

No, I did not.

Quoted 1 line

You did make that test didn't you?

Why do you need to be so aggressive ? It is possible that removing this option was a mistake, but there's *really* no need to be so aggressive. It only reduces the motivation to look at the problem...

The last time this subject came up on the ML tests seemed to indicate that the option was required by the Lua package.

I just presumed you have been following the mailing list postings.

Quoted 3 lines

So, trying to be constructive, how can I test that Lua still produces correct backtraces ? (I'm not a Lua programmer at all).

Pass an illegal value to a library routine or to a routine that you write. "Lua" refers to both a set of C libraries and to the interactive mode which uses those same libraries.

So testing with the interactive mode should be enough to see a backtrace.

Since the libraries (written in ANSI C) are quite close to the interactive environment (usually only one or two calls away) you will not see a very long backtrace unless you write a function or two of your own, but you will get one.

bye = function() return os.exit("String Value") end bye()

I.E:
mszick at wolf466:~$ lua
Lua 5.1.4 Copyright (C) 1994-2008 Lua.org, PUC-Rio

Quoted 1 line

os.exit("String Value")

stdin:1: bad argument #1 to 'exit' (number expected, got string) stack traceback:
[C]: in function 'exit'
stdin:1: in main chunk
[C]: ?

The "[C]" just means that is a hunk of compiled C code, for which the language does not have any more information to give you about it.

And: os.exit(0) should get you back out of the interactive Lua (or ctrl-D).

Mike

Quoted 3 lines

Thanks,

Thomas

_______________________________________________
buildroot mailing list
buildroot at busybox.net
http://lists.busybox.net/mailman/listinfo/buildroot

mszick at wolf466:~$ lua
Lua 5.1.4 Copyright (C) 1994-2008 Lua.org, PUC-Rio

Quoted 2 lines

bye = function() return os.exit("String Value") end bye()

stdin:1: bad argument #1 to 'exit' (number expected, got string) stack traceback:
[C]: in function 'exit'
stdin:1: in function 'bye'
stdin:1: in main chunk
[C]: ?

Hello Michael,

Quoted 10 lines · On Thu, 29 Jul 2010 07:18:56 -0500 "Michael S. Zick" <[email protected]>…

mszick at wolf466:~$ lua
Lua 5.1.4 Copyright (C) 1994-2008 Lua.org, PUC-Rio

Quoted 2 lines

bye = function() return os.exit("String Value") end bye()

stdin:1: bad argument #1 to 'exit' (number expected, got string) stack traceback:
[C]: in function 'exit'
stdin:1: in function 'bye'
stdin:1: in main chunk
[C]: ?

With the current Buildroot git (i.e, without SJLJ exceptions), I built an ARM system with Lua, and made your test under Qemu. I had the exact same output, with the exact same stack trace.

Are more complicated cases likely to make a difference ?

Regards,

Thomas

Hello Michael,

Quoted 10 lines · On Thu, 29 Jul 2010 07:18:56 -0500 "Michael S. Zick" <[email protected]>…

mszick at wolf466:~$ lua
Lua 5.1.4 Copyright (C) 1994-2008 Lua.org, PUC-Rio

Quoted 2 lines

bye = function() return os.exit("String Value") end bye()

stdin:1: bad argument #1 to 'exit' (number expected, got string) stack traceback:
[C]: in function 'exit'
stdin:1: in function 'bye'
stdin:1: in main chunk
[C]: ?

With the current Buildroot git (i.e, without SJLJ exceptions), I built an ARM system with Lua, and made your test under Qemu. I had the exact same output, with the exact same stack trace.

Are more complicated cases likely to make a difference ?

Regards,

Thomas

Quoted 3 lines

Why do you need to be so aggressive ? It is possible that removing this option was a mistake, but there's *really* no need to be so aggressive. It only reduces the motivation to look at the problem...

The last time this subject came up on the ML tests seemed to indicate that the option was required by the Lua package.

I just presumed you have been following the mailing list postings.

Pass an illegal value to a library routine or to a routine that you write. "Lua" refers to both a set of C libraries and to the interactive mode which uses those same libraries.

So testing with the interactive mode should be enough to see a backtrace.

Since the libraries (written in ANSI C) are quite close to the interactive environment (usually only one or two calls away) you will not see a very long backtrace unless you write a function or two of your own, but you will get one.

Quoted 1 line

Quoted 1 line

Quoted 1 line

Quoted 1 line

"Thomas" == Thomas Petazzoni <[email protected]> writes:

Quoted 4 lines · On Thu, 29 Jul 2010 06:59:36 -0500 "Michael S. Zick" <[email protected]>…

Quoted 3 lines

Why do you need to be so aggressive ? It is possible that removing this option was a mistake, but there's *really* no need to be so aggressive. It only reduces the motivation to look at the problem...

Quoted 4 lines

The last time this subject came up on the ML tests seemed to indicate that the option was required by the Lua package.

I just presumed you have been following the mailing list postings.

Sorry, I couldn't find the postings you're refering to. Do you have a pointer ?

Quoted 2 lines

Sorry, I couldn't find the postings you're refering to. Do you have a pointer ?

No I don't have a pointer handy.

Look for one of the threads that you terminated discussion with a:

"If you don't like it, you can hire commercial support for it from my firm."

It may even be the same thread where the ethics of using an "Open Source Project" as a "loss leader" to gain commercial customers was discussed.

Quoted 1 line

Quoted 1 line

Quoted 1 line

Quoted 1 line

"Thomas" == Thomas Petazzoni <[email protected]> writes: