qemu-devel
[Top][All Lists]
Advanced

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

Re: [Qemu-devel] [PATCH v4 08/33] tcg-aarch64: Introduce tcg_fmt_Rdnm an


From: Richard Henderson
Subject: Re: [Qemu-devel] [PATCH v4 08/33] tcg-aarch64: Introduce tcg_fmt_Rdnm and tcg_fmt_Rdnm_lsl
Date: Tue, 17 Sep 2013 07:54:00 -0700
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:17.0) Gecko/20130805 Thunderbird/17.0.8

On 09/17/2013 01:23 AM, Claudio Fontana wrote:
>> It would have been handy if ARM had officially assigned identifiers to the
>> formats, like Power, S390, and ia64 do.  Then one can build in the format ids
>> into both the function and enumeration names and use the preprocessor for
>> typechecking (c.f. the tcg_out_insn macro in  tcg/s390/tcg-target.c).
> 
> No need to do force explicit typechecking like that.
> That kind of use of the preprocessor really hurts.

Why do you believe this?  Have you browsed through the s390 backend?
I think it's a remarkably clean solution -- one we ought to have used
in the ia64 backend, which has even more format codes.


>> Therefore I think adding LSR, ASR and ROR shifts is both a waste of time and
>> bloats the backend.
> 
> I agree, lets just keep only left shift and right shift.
> Distinguishing the two costs one comparison per call, I think we can survive 
> it.

I don't understand you at all.

You agree that removing the unused shift from cmp is sensible.
You agree that not adding unused asr/ror shifts is sensible.
But you insist that the unused lsr shift should be retained?

You complain about wasting Y space in my positioning of comments.
But you insist on wasting X space, and allowing the possibility of
mismatch, by requiring format names to be duplicated?


r~



reply via email to

[Prev in Thread] Current Thread [Next in Thread]