--- Additional Comments From kris dot van dot hees at oracle dot com
2008-05-29 20:54 ---
Patch posted to the mailing list:
http://sourceware.org/ml/binutils/2008-05/msg00248.html
2008-05-29 Kris Van Hees <[EMAIL PROTECTED]>
* target-reloc.h (relocate_for_relocatable): Fix ne
--- Additional Comments From jevinsweval at gmail dot com 2008-05-29 21:39
---
Created an attachment (id=2766)
--> (http://sourceware.org/bugzilla/attachment.cgi?id=2766&action=view)
reduced test case
test with:
avr-gcc -c -o onewire.o -Os onewire.i
avr-gcc -Wl,--relax -o onewire.elf o
--- Additional Comments From jevinsweval at gmail dot com 2008-05-29 21:42
---
I've attached a reduced test case made with multidelta. It got minimized to 11
meaningful lines. It was run through the preprocessor before minimization so the
actual code is hard to read but I doubt that real
--- Additional Comments From hutchinsonandy at aim dot com 2008-05-29
22:38 ---
I applied this patch to Binutils-2.18.50
I ran full avr gcc testsuite and there were no regressions. Importantly, it
removed all the failures caused by the warning message removed by this patch.
So its get
--- Additional Comments From ian at airs dot com 2008-05-30 04:09 ---
This is now reportedly fixed in the current source code.
--
What|Removed |Added
Status|ASSIG
--- Additional Comments From ian at airs dot com 2008-05-30 04:18 ---
This is a known bug when using old versions of strip on x86_64. It happens
because the default page size for gold for x86_64 is 0x1000. The default page
size for old versions of the GNU binutils for x86_64 was 0x1
