Skip to content

R
riscv-gcc-1
Commit 6d0c87b5 by Uros Bizjak
Options

* ChangeLog: Fix whitespace. 

From-SVN: r180039
parent 89e4bd80
Showing with 19 additions and 30 deletions
gcc/ChangeLog
...@@ -50,8 +50,7 @@ ...@@ -50,8 +50,7 @@
tstsi_t_zero_extract_subreg_xor_big): New insns. tstsi_t_zero_extract_subreg_xor_big): New insns.
(*movsicc_t_false, *movsicc_t_true): Replace space with tab in (*movsicc_t_false, *movsicc_t_true): Replace space with tab in
asm output. asm output.
(*andsi_compact): Reorder alternatives so that K08 is considered (*andsi_compact): Reorder alternatives so that K08 is considered first.
first.
2011-10-14 Eric Botcazou <ebotcazou@adacore.com> 2011-10-14 Eric Botcazou <ebotcazou@adacore.com>
...@@ -74,8 +73,7 @@ ...@@ -74,8 +73,7 @@
vec_widen_smult_lo_v8hi, vec_widen_umult_hi_v8hi, vec_widen_smult_lo_v8hi, vec_widen_umult_hi_v8hi,
vec_widen_umult_lo_v8hi): Macroize using VI2_AVX2 vec_widen_umult_lo_v8hi): Macroize using VI2_AVX2
mode iterator and any_extend code iterator. mode iterator and any_extend code iterator.
(vec_widen_<s>mult_hi_v8si, vec_widen_<s>mult_lo_v8si): New (vec_widen_<s>mult_hi_v8si, vec_widen_<s>mult_lo_v8si): New expanders.
expanders.
(vec_widen_smult_hi_v4si, vec_widen_smult_lo_v4si): Enable (vec_widen_smult_hi_v4si, vec_widen_smult_lo_v4si): Enable
also for TARGET_SSE4_1 using pmuldq insn. also for TARGET_SSE4_1 using pmuldq insn.
(sdot_prodv8hi): Macroize using VI2_AVX2 iterator. (sdot_prodv8hi): Macroize using VI2_AVX2 iterator.
...@@ -96,8 +94,7 @@ ...@@ -96,8 +94,7 @@
2011-10-14 Jakub Jelinek <jakub@redhat.com> 2011-10-14 Jakub Jelinek <jakub@redhat.com>
* config/i386/sse.md (vec_interleave_high<mode>, * config/i386/sse.md (vec_interleave_high<mode>,
vec_interleave_low<mode>): Add AVX2 expanders for VI_256 vec_interleave_low<mode>): Add AVX2 expanders for VI_256 modes.
modes.
* config/i386/i386.c (expand_vec_perm_interleave3): New function. * config/i386/i386.c (expand_vec_perm_interleave3): New function.
(ix86_expand_vec_perm_builtin_1): Call it. (ix86_expand_vec_perm_builtin_1): Call it.
...@@ -127,8 +124,7 @@ ...@@ -127,8 +124,7 @@
* config/i386/sse.md (neg<mode>2): Use VI_AVX2 iterator instead * config/i386/sse.md (neg<mode>2): Use VI_AVX2 iterator instead
of VI_128. of VI_128.
* config/i386/sse.md (mulv2di3): Macroize using VI8_AVX2 * config/i386/sse.md (mulv2di3): Macroize using VI8_AVX2 iterator.
iterator.
(ashl<mode>3): Use VI248_AVX2 iterator instead of VI248_128. (ashl<mode>3): Use VI248_AVX2 iterator instead of VI248_128.
Use <sseinsnmode> instead of TI in mode attr. Use <sseinsnmode> instead of TI in mode attr.
...@@ -197,9 +193,8 @@ ...@@ -197,9 +193,8 @@
2011-10-14 Jakub Jelinek <jakub@redhat.com> 2011-10-14 Jakub Jelinek <jakub@redhat.com>
* gimple.c (walk_stmt_load_store_addr_ops): Call visit_addr * gimple.c (walk_stmt_load_store_addr_ops): Call visit_addr also
also on COND_EXPR/VEC_COND_EXPR comparison operands if they are on COND_EXPR/VEC_COND_EXPR comparison operands if they are ADDR_EXPRs.
ADDR_EXPRs.
2011-10-14 Richard Guenther <rguenther@suse.de> 2011-10-14 Richard Guenther <rguenther@suse.de>
...@@ -232,8 +227,7 @@ ...@@ -232,8 +227,7 @@
(vec_unpacks_float_hi_v8hi, vec_unpacks_float_lo_v8hi, (vec_unpacks_float_hi_v8hi, vec_unpacks_float_lo_v8hi,
vec_unpacku_float_hi_v8hi, vec_unpacku_float_lo_v8hi): Macroize vec_unpacku_float_hi_v8hi, vec_unpacku_float_lo_v8hi): Macroize
using VI2_AVX2 iterator. using VI2_AVX2 iterator.
(vec_unpacku_float_hi_v8si, vec_unpacku_float_lo_v8si): New (vec_unpacku_float_hi_v8si, vec_unpacku_float_lo_v8si): New expanders.
expanders.
2011-10-13 David S. Miller <davem@davemloft.net> 2011-10-13 David S. Miller <davem@davemloft.net>
...@@ -276,15 +270,13 @@ ...@@ -276,15 +270,13 @@
2011-10-13 Jakub Jelinek <jakub@redhat.com> 2011-10-13 Jakub Jelinek <jakub@redhat.com>
* config/i386/sse.md (vec_set<mode>): Change V_128 iterator mode * config/i386/sse.md (vec_set<mode>): Change V_128 iterator mode to V.
to V.
2011-10-13 Jakub Jelinek <jakub@redhat.com> 2011-10-13 Jakub Jelinek <jakub@redhat.com>
Richard Guenther <rguenther@suse.de> Richard Guenther <rguenther@suse.de>
* tree-ssa.c (maybe_optimize_var): Drop TREE_ADDRESSABLE * tree-ssa.c (maybe_optimize_var): Drop TREE_ADDRESSABLE from vector
from vector or complex vars even if their DECL_UID is in not_reg_needs or complex vars even if their DECL_UID is in not_reg_needs bitmap.
bitmap.
2011-10-13 Jakub Jelinek <jakub@redhat.com> 2011-10-13 Jakub Jelinek <jakub@redhat.com>
...@@ -356,7 +348,8 @@ ...@@ -356,7 +348,8 @@
(fmul8x16_vis, fmul8x16au_vis, fmul8sux16_vis, fmuld8sux16_vis): (fmul8x16_vis, fmul8x16au_vis, fmul8sux16_vis, fmuld8sux16_vis):
Reimplement as unspecs and remove inaccurate comments. Reimplement as unspecs and remove inaccurate comments.
(vis3_shift_patname): New code attr. (vis3_shift_patname): New code attr.
(<vis3_shift_insn><vbits>_vis): Rename to "v<vis3_shift_patname><mode>3". (<vis3_shift_insn><vbits>_vis): Rename to
"v<vis3_shift_patname><mode>3".
(vis3_addsub_ss_patname): New code attr. (vis3_addsub_ss_patname): New code attr.
(<vis3_addsub_ss_insn><vbits>_vis): Rename to (<vis3_addsub_ss_insn><vbits>_vis): Rename to
"<vis3_addsub_ss_patname><mode>3". "<vis3_addsub_ss_patname><mode>3".
...@@ -372,14 +365,12 @@ ...@@ -372,14 +365,12 @@
* config/i386/sse.md (vec_unpacks_lo_<mode>, * config/i386/sse.md (vec_unpacks_lo_<mode>,
vec_unpacks_hi_<mode>, vec_unpacku_lo_<mode>, vec_unpacks_hi_<mode>, vec_unpacku_lo_<mode>,
vec_unpacku_hi_<mode>): Change VI124_128 mode to vec_unpacku_hi_<mode>): Change VI124_128 mode to VI124_AVX2.
VI124_AVX2.
* config/i386/i386.c (ix86_expand_sse_unpack): Handle * config/i386/i386.c (ix86_expand_sse_unpack): Handle
V32QImode, V16HImode and V8SImode for TARGET_AVX2. V32QImode, V16HImode and V8SImode for TARGET_AVX2.
* config/i386/sse.md (vec_avx2): New mode_attr. * config/i386/sse.md (vec_avx2): New mode_attr.
(mulv16qi3): Macroize to cover also mulv32qi3 for (mulv16qi3): Macroize to cover also mulv32qi3 for TARGET_AVX2 into ...
TARGET_AVX2 into ...
(mul<mode>3): ... this. (mul<mode>3): ... this.
* config/i386/i386.md (UNSPEC_VPERMDI): Remove. * config/i386/i386.md (UNSPEC_VPERMDI): Remove.
...@@ -426,8 +417,7 @@ ...@@ -426,8 +417,7 @@
2011-10-12 Kai Tietz <ktietz@redhat.com> 2011-10-12 Kai Tietz <ktietz@redhat.com>
* config/i386/i386.md (simple_return): Disable * config/i386/i386.md (simple_return): Disable if TARGET_SEH is active.
if TARGET_SEH is active.
2011-10-12 Steve Ellcey <sje@cup.hp.com> 2011-10-12 Steve Ellcey <sje@cup.hp.com>
...@@ -638,7 +628,8 @@ ...@@ -638,7 +628,8 @@
2011-10-11 Georg-Johann Lay <avr@gjlay.de> 2011-10-11 Georg-Johann Lay <avr@gjlay.de>
* config/avr/avr-protos.h (avr_mode_code_base_reg_class): New prototype. * config/avr/avr-protos.h (avr_mode_code_base_reg_class): New
prototype.
(avr_regno_mode_code_ok_for_base_p): New prototype. (avr_regno_mode_code_ok_for_base_p): New prototype.
* config/avr/avr.h (BASE_REG_CLASS): Remove. * config/avr/avr.h (BASE_REG_CLASS): Remove.
(REGNO_OK_FOR_BASE_P): Remove. (REGNO_OK_FOR_BASE_P): Remove.
...@@ -708,8 +699,7 @@ ...@@ -708,8 +699,7 @@
2011-10-10 Georg-Johann Lay <avr@gjlay.de> 2011-10-10 Georg-Johann Lay <avr@gjlay.de>
* config/avr/avr.c (avr_option_override): Set * config/avr/avr.c (avr_option_override): Set
flag_omit_frame_pointer to 0 if frame pointer is needed for flag_omit_frame_pointer to 0 if frame pointer is needed for unwinding.
unwinding.
2011-10-10 Uros Bizjak <ubizjak@gmail.com> 2011-10-10 Uros Bizjak <ubizjak@gmail.com>
...@@ -854,8 +844,7 @@ ...@@ -854,8 +844,7 @@
(get_sched_window): Adjust for above field and SCHED_* macro changes. (get_sched_window): Adjust for above field and SCHED_* macro changes.
Use NODE_ASAP instead of SCHED_ASAP. Use NODE_ASAP instead of SCHED_ASAP.
(try_scheduling_node_in_cycle): Remove node parameter. Update (try_scheduling_node_in_cycle): Remove node parameter. Update
call to ps_add_node_check_conflicts. Pass ids to the SCHED_* call to ps_add_node_check_conflicts. Pass ids to the SCHED_* macros.
macros.
(sms_schedule_by_order): Update call to try_scheduling_node_in_cycle. (sms_schedule_by_order): Update call to try_scheduling_node_in_cycle.
(ps_insert_empty_row): Adjust for above field changes. (ps_insert_empty_row): Adjust for above field changes.
(compute_split_row): Use ids rather than nodes. (compute_split_row): Use ids rather than nodes.
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment