[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] clang compilation

>>> On 16.08.13 at 10:54, Tim Deegan <tim@xxxxxxx> wrote:
> At 09:48 +0100 on 16 Aug (1376646500), Jan Beulich wrote:
>> Yet with all of this I wonder what kind of broken assembler they're
>> using - deferring the operand size from register operands should
>> work consistently for all instructions or none, i.e. needing explicit
>> suffixes on mul but not any of the other .
> It's not just mul (see 794d4b9e85047aacfe23b852d3a03a8eff920aec in the
> original clang series) but it is certainly odd that it's inconsistent
> Presumably in some cases the choice is constrained by some other choice
> the optimizer has already made.

No, from my pov that other change has nothing to do with operand
size deduction. Especially the missing 's' suffixes on the fi...
instructions were an outright bug.

> Maybe one day I'll have the time to audit all the inline asm for
> implicit operand sizes. :)

Uglifying the code especially for those - like me - who come from
the Intel syntax world, which doesn't know suffixes except for
operands which have no way of expressing their size in non-
ambiguous ways.


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.