Re: Potential Mono removal from 32-bit big endian PowerPC in next upload
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA256
On 26/08/15 22:14, Lennart Sorensen wrote:
> Patch I did, which worked this well on the first try, but I won't
> say it is perfect for sure (and I suspect if it makes sense, then
> ppc64 could add the instructions to their .md file too).
>
> diff --git a/mono/mini/cpu-ppc.md b/mono/mini/cpu-ppc.md index
> e6baf91..8109c23 100644 --- a/mono/mini/cpu-ppc.md +++
> b/mono/mini/cpu-ppc.md @@ -128,6 +128,8 @@ storer8_memindex: dest:b
> src1:i src2:i len:4 loadu4_mem: dest:i len:8 move: dest:i src1:i
> len:4 fmove: dest:f src1:f len:4 +move_f_to_i4: dest:i src1:f
> len:8 +move_i4_to_f: dest:f src1:i len:8 add_imm: dest:i src1:i
> len:4 sub_imm: dest:i src1:i len:4 mul_imm: dest:i src1:i len:4
> diff --git a/mono/mini/mini-ppc.c b/mono/mini/mini-ppc.c index
> 8748c93..4d8fc14 100644 --- a/mono/mini/mini-ppc.c +++
> b/mono/mini/mini-ppc.c @@ -3745,6 +3745,14 @@
> mono_arch_output_basic_block (MonoCompile *cfg, MonoBasicBlock
> *bb) if (ins->dreg != ins->sreg1) ppc_fmr (code, ins->dreg,
> ins->sreg1); break; + case OP_MOVE_F_TO_I4: +
> ppc_stfs (code, ins->sreg1, -4, ppc_r1); +
> ppc_ldptr (code, ins->dreg, -4, ppc_r1); +
> break; + case OP_MOVE_I4_TO_F: +
> ppc_stw (code, ins->sreg1, -4, ppc_r1); +
> ppc_lfs (code, ins->dreg, -4, ppc_r1); +
> break; case OP_FCONV_TO_R4: ppc_frsp (code, ins->dreg,
> ins->sreg1); break;
Okay. At this point, I've tried testing on a different box I was
generously offered access to, and it works as well as you indicate -
so my suspicion is there is Something Weird(tm) on partch.debian.org.
I intend to upload the next upstream bugfix version of Mono early next
week, with a katamari if your patches & the patches in my ppc32 branch
- - I don't know if the buildd will lean closer to the "fine" or "on
crack" side of things, but I don't intend to remove powerpc from the
arch list with this upload.
Would you be able to file GH pull requests with your fixes? They're
good stuff (and in some cases generally useful for BE) so it'd help to
get those merged upstream for all users
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v2
iQEcBAEBCAAGBQJV4BSeAAoJEMkPnLkOH60MBb8H+gIvdwtVBM1TQ/4gLD0Zyxf3
e5x7IlaolaE5QOrFgwBLObU/HsS1O216tdm5JD9CX1wAtSZBeuqZDOWBZpL9YfEJ
1JuV4WtBKSR4bdqD6196Hw+X8DZv7NFPDLWR4DGaAVgbfEv0I7r3p3lSkf+QLutH
IlEjeEZvzgVKR2MyBPkLpGp1sfuIo3lQGKR9hUDLjuQs+qZamspQ2fwGxD41oYQ9
/XIbwo8pHEHYuPHloK62hu1KJZ6IYU+4tDC4XKd+GjBBx7zmTCHGiww/AwlhWJE4
067A7WScmnACcZU0vQ42Hw8XGc478Izr14PWJ5yt8uLHFd1UBXLF5bQKNQQu+aI=
=n6h8
-----END PGP SIGNATURE-----
Reply to: