[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastr
From: |
Richard Henderson |
Subject: |
Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions |
Date: |
Sat, 25 Jan 2020 08:06:40 -1000 |
User-agent: |
Mozilla/5.0 (X11; Linux x86_64; rv:68.0) Gecko/20100101 Thunderbird/68.4.1 |
On 1/25/20 12:22 AM, Aleksandar Markovic wrote:
>
>
> On Saturday, January 25, 2020, Richard Henderson <address@hidden
> <mailto:address@hidden>> wrote:
>
> On 1/24/20 1:38 PM, Aleksandar Markovic wrote:
> > On Friday, January 24, 2020, Richard Henderson
> <address@hidden <mailto:address@hidden>
> > <mailto:address@hidden
> <mailto:address@hidden>>> wrote:
> > The thing I'm concerned about here is any future maintenance of this
> file. One
> > would surely prefer to edit the original decodetree input than this
> output.
> >
> >
> > Here is the deal: This dissasembler is meant to reflect the
> documentation of a
> > particular ISA, and as the documentation largely stays constant (except
> adding
> > some insignificant errata), the disassembler will stay virtually
> constant, we
> > wouldn't like even to touch it, and we like it this way.
>
> No, this is neither right nor proper.
>
> To review the code in this form is significantly harder than in its
> decodetree
> form. That is in fact the whole point of the decodetree form: otherwise
> we'd
> still be writing these sorts of parsers by hand.
>
> While there's no license on this new file (another problem), if as
> assumed this
> is GPL 2+, then you are in violation of the GPL. From section 3:
>
> # The source code for a work means the preferred form of
> # the work for making modifications to it.
>
> You cannot with a straight face claim that the generated c is the
> preferred
> form for making modifications.
>
> Finally, suppose we improve decodetree.py, so that it produces code with
> which
> the compiler produces better code, for some metric of better. We would
> want
> this disassembler to benefit as well.
>
>
> I think you got some things upside-down. A tool developent and usage should be
> driven by the needs of users of a tool, and not dictated by the author. Users
> should be free to use the tool in any way they seem suitable, including its
> modification.
That's exactly right, provided that by "tool" you mean QEMU and by "users" you
mean everyone downstream of us.
Let me state for the record that I object to this generated file being merged
in this form, for the reasons I stated above.
r~
- [PATCH v4 2/7] target/mips: Add support for MIPS<32|64>R6 CRC32 ISA, (continued)
- [PATCH v4 2/7] target/mips: Add support for MIPS<32|64>R6 CRC32 ISA, Aleksandar Markovic, 2020/01/24
- [PATCH v4 4/7] target/mips: Add implementation of GINVT instruction, Aleksandar Markovic, 2020/01/24
- [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Aleksandar Markovic, 2020/01/24
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Richard Henderson, 2020/01/24
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Aleksandar Markovic, 2020/01/24
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Richard Henderson, 2020/01/24
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Aleksandar Markovic, 2020/01/24
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Richard Henderson, 2020/01/25
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions, Aleksandar Markovic, 2020/01/25
- Re: [PATCH v4 6/7] disas: mips: Add micromips R6 disassembler - infrastructure and 16-bit instructions,
Richard Henderson <=
[PATCH v4 7/7] disas: mips: Add micromips R6 disassembler - 32-bit instructions, Aleksandar Markovic, 2020/01/24
Re: [PATCH v4 0/7] target/mips: Misc MIPS fixes and improvements for 5.0, no-reply, 2020/01/24
Re: [PATCH v4 0/7] target/mips: Misc MIPS fixes and improvements for 5.0, no-reply, 2020/01/24