> - providing an emulation host for any target ARCH other than x86_32, since

tangent, since we mention qemu in a tcc context (just fyi, I don't expect any 
source code diffs):

I wanted to mention this http://landley.net/qcc/
re-using the code in qemu's TCG to support more backends
afaict this means emitting TCG IR (from tcc) for the TCG code (call some qemu 
lib from tcc) to then emit to machine code

the qemu upstream is a bit nasty in terms of build dependencies, I wanted to 
ask them for a public TCG library .h file because I don't want to deal with 
their big collection of build-system code

this could also help with cross-compilation: the proposed "qcc" emits TCG IR to 
qemu's TCG, which does the rest - we only need to do the "C to TCG IR" step 
correctly
this additional processing step will probably decrease the compilation speed
we can also go even bigger: targeting SPIR-V instead - which has a thorough 
spec, unlike TCG IR

I'd say that neither fit into tcc, it wouldn't be tiny anymore
We can look at the qemu TCG backends (TCG IR->machine code) regardless - 
remember that qemu has common ancestry with tcc - with n years of adding 
complexity in hopes of better throughput

Attachment: signature.asc
Description: PGP signature

_______________________________________________
Tinycc-devel mailing list
[email protected]
https://lists.nongnu.org/mailman/listinfo/tinycc-devel

Reply via email to