This is the mail archive of the mailing list for the binutils project.

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: ENTER/BOUND operands order.

Dnia 2014-01-16 09:20 Jan Beulich napisaÅ(a):

>>>> On 15.01.14 at 22:31, "Slawomir Wojtasiak" 
<> wrote:
>> Dnia 2014-01-07 11:09 Jan Beulich napisaÅ(a):
>>>In fact, the amount of inconsistencies appears to continuously
>>>increase in particular with new SIMD additions, recently in bogus
>>>operand swapping requiring to use an incorrect operand order
>>>in Intel syntax:
>>>	vcvtsi2ss	xmm0, xmm0, {rn-sae}, eax
>>>	vcvtusi2ss	xmm0, xmm0, {rn-sae}, eax
>>>(where in fact the rounding mode specifier ought to be the
>>>last operand).
>> Exactly, GAS can be at cutting edge of the AT&T dialect, but I'm pretty 
>> that it should strictly follow Intel dialect syntax, so encoding in the 
>> example above is probably a bug. These instructions are relatively new, 
>> they probably will be fixed in near future. For instance I found a lot of 
>> broken SIMD/AVX instructions in GDB 7.1 last year and all of them have 
>> already fixed in the current version of the GNU debugger (and the fact 
>> they was shipped wasn't the problem), so we have to hope for the best ! 
>Except that fixing disassembler bugs (i.e. what affects gdb) of this
>kind is always possible, whereas fixing assembler bugs isn't: People
>may have written (wrong) code already that they expect to the
>assembler to accept if it once did. I therefore understand H.J.'s
>reservations here to a certain degree - but I nevertheless want the
>assembler to strive towards accepting valid code and rejecting
>broken stuff (the latter to reduce surprises when moving or reusing
>code between different assemblers).

Unfortunately I didn't check these instructions with corresponding version 
of GNU assembler, so I don't know if they were also broken or not. Anyway, 
of course I agree with you that such bug fixes are always problematic, 
nevertheless I think that it's the last chance to fix them without big 
consequences. Intel dialect is probably not so widely used as AT&T is in 
case of GAS and these instructions aren't very popular yet, so it should not 
make a big mess.

SÅawomir Wojtasiak

Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]