------- Comment #5 from hjl at lucon dot org 2006-05-05 13:01 -------
There are several problems with the current SEE implementation:
1. SEE uses NEXT_INSN/PREV_INSN to find adjacent insns to check if SEE
is safe. But with -g, NEXT_INSN/PREV_INSN may point to a NOTE:
(note:HI 17 14 18 2
("/net/gnu-13/export/gnu/src/gcc-see/gcc/gcc/testsuite/gcc.c-torture/execute/20010224-1.c")
18)
(insn:HI 18 17 19 2
/net/gnu-13/export/gnu/src/gcc-see/gcc/gcc/testsuite/gcc.c-torture/execute/20010224-1.c:18
(set (reg/v:SI 70 [ j ])
(sign_extend:SI (subreg:HI (reg:SI 72 [ start ]) 0))) 118 {extendhisi2}
(insn_list:REG_DEP_TRUE 12 (nil))
(expr_list:REG_DEAD (reg:SI 72 [ start ])
(nil)))
(note:HI 19 18 20 2
("/net/gnu-13/export/gnu/src/gcc-see/gcc/gcc/testsuite/gcc.c-torture/execute/20010224-1.c")
19)
(insn:HI 20 19 22 2
/net/gnu-13/export/gnu/src/gcc-see/gcc/gcc/testsuite/gcc.c-torture/execute/20010224-1.c:19
(set (reg:DI 73 [ j ])
(sign_extend:DI (reg/v:SI 70 [ j ]))) 115 {extendsidi2_rex64}
(insn_list:REG_DEP_TRUE 18 (nil))
(nil))
Those notes may be added between insns because -g. NEXT_INSN/PREV_INSN
won't get you the adjacent insn in this case.
2. Not all zero_extend patterns are available for x86/x86-64. For
example:
(insn 137 0 0 (set (reg:SI 60 [ prephitmp.115 ])
(zero_extend:SI (subreg:QI (reg:SI 60 [ prephitmp.115 ]) 0)))
-1 (nil)
(nil))
may not be used on x86/x86-64. i386.md has
(define_expand "zero_extendqisi2"
[(parallel
[(set (match_operand:SI 0 "register_operand" "")
(zero_extend:SI (match_operand:QI 1 "nonimmediate_operand" "")))
(clobber (reg:CC FLAGS_REG))])]
""
"")
This is case for all extensions for i386. For x86-64, only
zero_extendsidi2 won't clobber CC. But SEE doesn't provide a way for a backend
to deal with it.
3. When the original insns were
set (dest_extension_reg1) (sign_extend (source_extension_reg1))
set (dest_extension_reg2) (sign_extend (dest_extension_reg1))
We created
ref: set (dest_extension_reg1) (sign_extend (source_extension_reg1))
def_se: set (dest_extension_reg2) (sign_extend (dest_extension_reg1))
and
use_se: set (dest_extension_reg1) (sign_extend (dest_extension_reg1))
ref: set (dest_extension_reg2) (sign_extend (dest_extension_reg1))
When def merge failed, def_se was deleted. Now use_se had a deleted
ref.
Basically, SEE doesn't handle
(set (reg/v:SI 70 [ j ])
(sign_extend:SI (subreg:HI (reg:SI 72 [ start ]) 0)))
(set (reg:DI 73 [ j ])
(sign_extend:DI (reg/v:SI 70 [ j ])))
correctly.
4. SEE also failed to handle
set (dest_extension_reg1) (zero_extend (source_extension_reg1))
set (reg) (..dest_extension_reg1..)
set (dest_extension_reg2) (sign_extend (source_extension_reg1))
(insn:HI 28 26 30 2 x.c:1201 (set (reg:DI 534 [ mode ])
(zero_extend:DI (reg/v:SI 264 [ mode ]))) 111
{zero_extendsidi2_rex64}
(insn_list:REG_DEP_TRUE 11 (nil))
(nil))
(insn:HI 30 28 269 2 x.c:1201 (set (reg:QI 261 [ D.24257 ])
(mem/s/u:QI (plus:DI (reg:DI 534 [ mode ])
(symbol_ref:DI ("mode_class") [flags 0x40] <var_decl
0x2a98a42630 mode_class>)) [0 mode_class S1 A8])) 55 {*movqi_1}
(insn_list:REG_DEP_TRUE 28 (nil))
(nil))
(insn:HI 269 30 270 2 x.c:1273 (set (reg:DI 546)
(sign_extend:DI (reg/v:SI 264 [ mode ]))) 115
{extendsidi2_rex64} (nil)
(nil))
#3 and #4 may happen since SEE uses NEXT_INSN/PREV_INSN to check
the adjacent insn. When -g is used, SEE may see the note instead of
the real adjacent insn and reaches wrong conclusion. It may lead to
compiler crash or wrong code.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=27437
