Commit graph

16 commits

Author SHA1 Message Date
cinap_lenrek 2d59b15c39 5c/6c/8c/vc: import various changes from charles forsyth
- cover more cases that have no side effects
- ensure function has complex FNX
- pull operators out of OFUNC level
- rewrite OSTRUCT lhs to avoid all side-effects, use regalloc() instead of regret()
2015-10-06 06:20:01 +02:00
cinap_lenrek ec51877698 6c: remove 6c/vlrt.c file 2015-10-04 22:12:16 +02:00
cinap_lenrek bfd0cb0db4 cc/6c: fix return type of mixed asop expressions, preserve type for moves so fixed<->float conversions work correctly 2015-10-04 20:02:58 +02:00
cinap_lenrek f5f9ecdcfa 6c: MOVL xxx, r; MOVLQZX r, r -> MOVL xxx, r
eleminate MOVLQXZ instructions after MOVL as MOVL implicitely
zero extends the result.
2015-03-17 16:12:01 +01:00
cinap_lenrek 7cdc13accc 6c: fix missing memset in Bconv() 2015-03-01 05:33:41 +01:00
cinap_lenrek 0037c93433 6c: eleminate more MOV instructions
convert:

x = B || W
MOVxLZX a, r; MOVxQZX r, b -> MOVxQZX a, r; MOVQ r, b
MOVxLSX a, r; MOVxQSX r, r -> MOVxQSX a, r; MOVQ r, r

the MOVQ can then be eleminated by copy propagation.

improve subprop() by accepting other mov and lea
instructions as the source op.
2015-02-18 23:34:33 +01:00
cinap_lenrek 03feba8cc1 [125678kqv][cl]: fix sprint() and strcpy() buffer overflows 2015-02-17 22:13:35 +01:00
cinap_lenrek decc7ec518 6c/8c: eleminate moves by swaping source and destination operands in peephole pass 2014-09-24 20:45:16 +02:00
cinap_lenrek bc306a5a63 8c, 6c: generate enam.c file, just like 5c 2014-08-07 21:35:52 +02:00
Aram Hăvărneanu bf0d5c8abb 6a, 6c, 6l: fix copy propagation
Without an explicit signal for a truncation, copy propagation will
sometimes propagate a 32-bit truncation and end up overwriting uses of
the original 64-bit value.

This was independently discovered and fixed in Go. See:
	http://golang.org/issue/1315
	https://codereview.appspot.com/6002043/

Thanks Charles Forsyth for tips and advice.
2014-05-30 12:28:01 +02:00
cinap_lenrek 391198888a 8c, 6c: fix peephole bug for eleminating CMPL $0,R after shift
the shift instructions does not change the zero flag
when the shift count is 0, so we cannot remove the
compare instruction in this case.

this fixes oggdec under 386.
2014-04-28 22:53:50 +02:00
cinap_lenrek cde97a4d5f 6c, 8c: optimize away CMPL/CMPQ reg, $0 instruction in peephole pass
when the previous instruction sets the zero flag,
we can remove the CMPL/CMPQ instruction.
this removes compares for zero/non zero tests only.
it only looks at the previous non-nop instruction
to see if it sets our compare value register.
2014-03-29 19:44:04 +01:00
cinap_lenrek 393dfd1ced 8c, 6c: fix mulgen botch error for handling multiplication by zero constant 2014-03-21 19:05:17 +01:00
cinap_lenrek 03ae4fa162 6c: fix 32bit pointer truncation (from patch/6c-sugen-types)
1. Go group spotted that a slightly-obscured pointer move was done by AMOVL not AMOVQ.
2. Inspecting the code further, I noticed that other pointer types were set to TLONG not TIND,
   causing similar truncation of pointers to 32 bits.
2013-02-28 19:23:54 +01:00
cinap_lenrek 3ba213a9d7 6c: extern register fix (import from patch/6c-extreg)
to make it easy to use normal libraries (such as libdraw, libsec, and libmp)
with the kernel, which uses extern register, don't stray into the external
register set when allocating values to registers.
2012-09-18 18:18:43 +02:00
cinap_lenrek 4f33c88a51 import updated compilers from sources 2012-07-30 19:11:16 +02:00