Binutils branch 2.20 (done!)

[Jie Zhang]

Mike Frysinger wrote:
> On Saturday 05 September 2009 08:39:26 Jie Zhang wrote:
>> Alan Modra wrote:
>>> On Fri, Sep 04, 2009 at 12:00:55PM +0200, Tristan Gingold wrote:
>>>> I have just created the 2.20 branch and updated the version on HEAD.
>>>> I will try to do the first prerelease soon.
>>> Lots of testsuite failures at the moment.  It would be nice if target
>>> maintainers would give these some attention.
>>>
>>> bfin-elf/log0:FAIL: flow2
>>> bfin-elf/log0:FAIL: loop
>>> bfin-elf/log0:FAIL: loop2
>>> bfin-elf/log0:FAIL: loop3
>> I don't see these 4 FAILs here.
>>
>>> bfin-elf/log0:FAIL: ld-elf/64ksec-r
>>> bfin-elf/log0:FAIL: ld-elf/64ksec
>> These two FAIL is because Blackfin as does not accept "SYMBOL = VALUE".
>> Blackfin has an algebra assembly syntax. Assigning value to a register
>> has the form:
>>
>> "REG = VALUE"
>>
>> If there is typo of REG, this will become "SYMBOL = VALUE". That error
>> is not easy to notice. So I think it's better to let Blackfin gas reject
>> "SYMBOL = VALUE" to detect such error. .set can be used if user want to
>> set value to a symbol.
>>
>> How about the attached patch?
> 
> i recall some tests semi-recently being changed from ".set" to "=" on purpose.  
> might want to double check the history.  maybe Alan remembers better than i 
> though ...

Thanks for reminding me of that. I think you meant this:

http://sourceware.org/ml/binutils/2009-04/msg00218.html

I have modify my patch. So it only uses ".set" instead of "=" for 
bfin-*-* targets. Tested for bfin-elf and alpha-linux.


Jie

-------------- next part --------------
A non-text attachment was scrubbed...
Name: ld-testsuite-sec64k-set-symbole-value-2.diff
Type: text/x-patch
Size: 1948 bytes
Desc: not available
URL: <https://sourceware.org/pipermail/binutils/attachments/20090905/f236bb04/attachment.bin>