[Top][All Lists]

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

Re: [Tinycc-devel] Re: C67 Support

From: Fabrice Bellard
Subject: Re: [Tinycc-devel] Re: C67 Support
Date: Sun, 16 Mar 2003 14:20:16 +0100
User-agent: Mozilla/5.0 (X11; U; Linux i586; en-US; rv:1.1) Gecko/20020828

Peter "Firefly" Lund wrote:

I don't know if elf even makes sense for C67 - I still have very little
experience with it.  (and I haven't read the two manuals for it yet -
I'll have to print them out at Uni)

ELF is very simple if no relocation are done. All we need is to define a correspondant machine id for TMS320C6xx.

elf generation for c67 should be very easy
especially if no DLLs are generated. Ideally, target endianness should
also be selectable (what about big endian x86 ? :-)).

I would /love/ that.  Did you see my idea for having selectable integer
and pointer sizes?

Ok for pointer and longs (it is needed for 64 bit CPUs). For ints I am not sure it is useful and it would be more complicated.

Do you agree that only compile time selection is needed for that ?

Note that my new project is a high performance portable Linux CPU user
space emulator.

Interesting.  What is wrong with bochs and valgrind? ;)

Bochs is incredibly slow and written in C++. valgrind is not portable and does not emulate full user mode (in particular full selector support and vm86 mode are missing). My CPU emulator is fast (it is a JIT as valgrind) and portable (I use some GCC tricks to avoid writing any code generator). The main limitation is that only user space Linux is emulated.

Emulators are fun, but I think the world needs a usable Linux debugger
more (which is what I'm working on).

Interesting. A good emulator will be useful if you want higher level debugging (think of execution trace recording/single step backward).


reply via email to

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