antcc/irdump.c, branch trunkA little C compilerhttps://git.lemon.rip/antcc/atom?h=trunk2025-10-19T06:09:09+00:00Organize source files into directories2025-10-19T06:09:09+00:00lemonlsof@mailbox.org2025-10-19T06:09:09+00:00urn:sha1:dea8fd171acb54b6d9685422d5e391fb55074008misc2025-10-15T08:45:12+00:00lemonlsof@mailbox.org2025-10-15T08:45:12+00:00urn:sha1:437efda31319573dcf67337d1cc6e4a23a6967a4regalloc: start implementing linear scan2025-09-08T20:05:33+00:00lemonlsof@mailbox.org2025-09-08T20:05:33+00:00urn:sha1:e043811980db560fc2507bb53b644e54c80527dcadd initializers (only static for initialier list rn)2023-06-29T07:59:30+00:00lemonlsof@mailbox.org2023-06-29T07:59:30+00:00urn:sha1:f453b313f62ba42d748f00628be7b3750c797c86
and other fixes
backend: fix mem2reg & regalloc2023-06-25T22:29:07+00:00lemonlsof@mailbox.org2023-06-25T22:29:07+00:00urn:sha1:b94fe89c9ddfcb85dcddebfd218fa7f00b8e6608
they were broken, especially for unstructured control flow. most
significant fix is to register allocator for temporaries that are used
before the first definition in the source order, e.g.:
@1:
%x = add %y, 1
b @3
@2
%y = ...
b @1
it's legal for %x to use %y there (assuming @2 dominates @1) but from
the point of view of the register allocator %y is defined and freed
and then used again, which broke things. the fix is to introduce phis
for this situation:
@1:
%y.1 = phi @2 %y
%x = add %y.1, 1
b @3
@2
%y = ...
b @1
then regalloc phi handling code makes it work
backend: fix regalloc to work with more complex dataflow2023-06-24T16:47:05+00:00lemonlsof@mailbox.org2023-06-24T16:47:05+00:00urn:sha1:19bbdfa3c7ae05f4694ce5e434d9855c6f2c3682
basically an allocation map at the beginning (in) and end (out) of each
block is kept and after the first allocation pass another pass is ran to
resolve allocation conflicts between each edge, plus another pass to
finish lowering phi functions.
also introduced `regset` and plenty of other miscellaneous fixes
change RMORE -> RADDR; use RXXX (RNONE) for special args,also undef2023-06-22T19:54:08+00:00lemonlsof@mailbox.org2023-06-22T19:54:08+00:00urn:sha1:d313c6e49bfb32ae24745e90eebe833da20efa1amem2reg: implement ssa construction; this breaks regalloc right now2023-06-21T21:52:23+00:00lemonlsof@mailbox.org2023-06-21T21:52:23+00:00urn:sha1:3f2221dfb9ab33b7ac44bbf822a78753a0357d25explicitly store predecessors in each block2023-06-21T10:32:32+00:00lemonlsof@mailbox.org2023-06-21T10:32:32+00:00urn:sha1:995fd23ecd5de710a6f587d29af2874b1fb4756ddon't lower params and args to registers in abi02023-06-20T09:25:51+00:00lemonlsof@mailbox.org2023-06-20T08:58:17+00:00urn:sha1:3b9071be52359388a4a1db6b48636f6c4934b9ad
it's better to do it later, currently in isel(), but perhaps in a
pre-isel abi1 pass