[PATCH] [RFT] m68k: enable HAVE_LD_DEAD_CODE_DATA_ELIMINATION

[PATCH] [RFT] m68k: enable HAVE_LD_DEAD_CODE_DATA_ELIMINATION

[Arnd Bergmann]

From: Arnd Bergmann <arnd@arndb.de>

While experimenting with CONFIG_TRIM_UNUSED_KSYMS across architectures,
I noticed that this unsurprisingly works best when combined with the
--gc-sections linker flag, which is not currently allowed on m68k.

Enabling it with this patch shows significant improvements
for kernel size, looking at m68k defconfig with gcc-10, I get

4005135 1374302 167108 5546545 54a231 vmlinux-normal
3916254 1378078 167108 5461440 5355c0 vmlinux+trim
4012933 1362514 164280 5539727 54878f vmlinux+gcsection
3797884 1334194 164640 5296718 50d24e vmlinux+gcsection+trim

There are probably bugs somewhere, but if someone is able to test this
patch and fix any issues that come up, would be a useful improvement
for memory-limited machines.

Link: https://lore.kernel.org/lkml/CAK8P3a05VZ9hSKRzVTxTn+1nf9E+gqebJWTj6N23nfm+ELHt9A@mail.gmail.com/
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
---
 arch/m68k/Kconfig | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/m68k/Kconfig b/arch/m68k/Kconfig
index c186733562b3..5f1aafa7b2e2 100644
--- a/arch/m68k/Kconfig
+++ b/arch/m68k/Kconfig
@@ -24,6 +24,7 @@ config M68K
 	select HAVE_DEBUG_BUGVERBOSE
 	select HAVE_FUTEX_CMPXCHG if MMU && FUTEX
 	select HAVE_IDE
+	select HAVE_LD_DEAD_CODE_DATA_ELIMINATION
 	select HAVE_MOD_ARCH_SPECIFIC
 	select HAVE_UID16
 	select MMU_GATHER_NO_RANGE if MMU
-- 
2.29.2

Re: [PATCH] [RFT] m68k: enable HAVE_LD_DEAD_CODE_DATA_ELIMINATION

[John Paul Adrian Glaubitz]

Hi Arnd!

On 2/25/21 12:14 PM, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> While experimenting with CONFIG_TRIM_UNUSED_KSYMS across architectures,
> I noticed that this unsurprisingly works best when combined with the
> --gc-sections linker flag, which is not currently allowed on m68k.
> 
> Enabling it with this patch shows significant improvements
> for kernel size, looking at m68k defconfig with gcc-10, I get
> 
> 4005135 1374302 167108 5546545 54a231 vmlinux-normal
> 3916254 1378078 167108 5461440 5355c0 vmlinux+trim
> 4012933 1362514 164280 5539727 54878f vmlinux+gcsection
> 3797884 1334194 164640 5296718 50d24e vmlinux+gcsection+trim
> 
> There are probably bugs somewhere, but if someone is able to test this
> patch and fix any issues that come up, would be a useful improvement
> for memory-limited machines.

Thanks, that sounds like a very helpful improvement. I'll test the patch next
week when I'm on vacation and report back if it works - unless someone beats
me to it ;-).

Adrian

-- 
 .''`.  John Paul Adrian Glaubitz
: :' :  Debian Developer - glaubitz@debian.org
`. `'   Freie Universitaet Berlin - glaubitz@physik.fu-berlin.de
  `-    GPG: 62FF 8A75 84E0 2956 9546  0006 7426 3B37 F5B5 F913

Re: [PATCH] [RFT] m68k: enable HAVE_LD_DEAD_CODE_DATA_ELIMINATION

[Kees Cook]

On Thu, Feb 25, 2021 at 12:14:45PM +0100, Arnd Bergmann wrote:
> From: Arnd Bergmann <arnd@arndb.de>
> 
> While experimenting with CONFIG_TRIM_UNUSED_KSYMS across architectures,
> I noticed that this unsurprisingly works best when combined with the
> --gc-sections linker flag, which is not currently allowed on m68k.
> 
> Enabling it with this patch shows significant improvements
> for kernel size, looking at m68k defconfig with gcc-10, I get
> 
> 4005135 1374302 167108 5546545 54a231 vmlinux-normal
> 3916254 1378078 167108 5461440 5355c0 vmlinux+trim
> 4012933 1362514 164280 5539727 54878f vmlinux+gcsection
> 3797884 1334194 164640 5296718 50d24e vmlinux+gcsection+trim
> 
> There are probably bugs somewhere, but if someone is able to test this
> patch and fix any issues that come up, would be a useful improvement
> for memory-limited machines.
> 
> Link: https://lore.kernel.org/lkml/CAK8P3a05VZ9hSKRzVTxTn+1nf9E+gqebJWTj6N23nfm+ELHt9A@mail.gmail.com/
> Signed-off-by: Arnd Bergmann <arnd@arndb.de>

Reviewed-by: Kees Cook <keescook@chromium.org>

-- 
Kees Cook