https://gcc.gnu.org/g:4e377ef17e9985ce6bad303b39a6975e20f7eb05
commit 4e377ef17e9985ce6bad303b39a6975e20f7eb05 Author: Michael Meissner <[email protected]> Date: Thu Jun 18 23:04:14 2026 -0400 Update ChangeLog.* Diff: --- gcc/ChangeLog.float | 69 ++++++++++++++++++++++++++++------------------------- 1 file changed, 36 insertions(+), 33 deletions(-) diff --git a/gcc/ChangeLog.float b/gcc/ChangeLog.float index 2b479e4c9c78..b69aa8b018aa 100644 --- a/gcc/ChangeLog.float +++ b/gcc/ChangeLog.float @@ -1,19 +1,8 @@ -==================== Branch work247-float, patch #211 ==================== - -Add BF to HF and HF to BF conversions - -2026-06-18 Michael Meissner <[email protected]> - -gcc/ - - * config/rs6000/float16.md (trunchfbf2): New insn. - (expandbfhf2): Likewise. - -==================== Branch work247-float, patch #210 ==================== +==================== Branch work247-float, patch #230 ==================== Add _Float16 and __bfloat16 tests. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/testsuite/ @@ -28,11 +17,11 @@ gcc/testsuite/ (check_ppc_bfloat16_runtime_available): Likewise. (is-effective-target): Add new _Float16 and __bfloat16 targets. -==================== Branch work247-float, patch #209 ==================== +==================== Branch work247-float, patch #229 ==================== Add --with-powerpc-float16 and --with-powerpc-float16-disable-warning. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -44,7 +33,7 @@ gcc/ (ISA_2_7_MASKS_SERVER): Likewise. (POWERPC_MASKS): Likewise. -==================== Branch work247-float, patch #208 ==================== +==================== Branch work247-float, patch #228 ==================== Optimize __bfloat16 scalar code. @@ -90,7 +79,7 @@ BFmode because GCC will not generate these patterns unless the user uses -Ofast. Without -Ofast, it will always convert BFmode into SFmode. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -113,11 +102,11 @@ gcc/ * config/rs6000/rs6000-protos.h (bfloat16_operation_as_v4sf): New declaration. -==================== Branch work247-float, patch #200 ==================== +==================== Branch work247-float, patch #227 ==================== Add 16-bit floating point vectorization. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -151,11 +140,11 @@ gcc/ (fp16_vectorization): New declaration. * config/rs6000/t-rs6000 (float16.o): Add build rules. -==================== Branch work247-float, patch #206 ==================== +==================== Branch work247-float, patch #226 ==================== Add BF/HF neg, abs operands and logical insns. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -176,11 +165,11 @@ gcc/ (boolc<mode>3): Likewise. (boolcc<mode>): Likewise. -==================== Branch work247-float, patch #205 ==================== +==================== Branch work247-float, patch #225 ==================== Add conversions between 16-bit floating point and other scalar modes. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -195,7 +184,7 @@ gcc/ (fix_trunc<FP16_HW:mode><GPR:mode>): Likewise. (fixuns_trunc<FP16_HW:mode><GPR:mode>2): Likewise. -==================== Branch work247-float, patch #204 ==================== +==================== Branch work247-float, patch #224 ==================== Add conversions between __bfloat16 and float/double. @@ -234,7 +223,7 @@ float/double value. GCC will generation. xscvdpsp 0,0 Convert float scalar to float memory format xvcvspbf16 0,0 Convert vector float to vector __bfloat16 -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -254,6 +243,8 @@ gcc/ (shift_bf_16bits): Likewise. (trunc<mode>bf): New insns to convert SFmode/DFmode to BFmode. (vsx_xscvdpspn_sf): Likewise. + (trunchfbf2): New insn to convert BFmode to HFmode. + (extendbfhf2): New insn to convert HFmode to BFmode. (cvt_fp16_to_v4sf_<mode): Likewise. (cvt_fp16_to_v4sf_<mode>_le): Likewise. (cvt_fp16_to_v4sf_<mode>_be): Likewise. @@ -266,14 +257,14 @@ gcc/ * config/rs6000/rs6000.cc (rs6000_init_hard_regno_mode_ok): Mark that we use VSX arithmetic support for V8BFmode if we are a power10 or later. -==================== Branch work247-float, patch #203 ==================== +==================== Branch work247-float, patch #223 ==================== Add conversions between _Float16 and float/double. This patch adds support to generate xscvhpdp and xscvdphp on Power9 systems and later, to convert between _Float16 and float scalar values. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -286,14 +277,14 @@ gcc/ * config/rs6000/rs6000.cc (rs6000_init_hard_regno_mode_ok): Mark that we use VSX arithmetic support for V8HFmode if we are a power9 or later. -==================== Branch work247-float, patch #202 ==================== +==================== Branch work247-float, patch #222 ==================== Add HF/BF emulation functions to libgcc. This patch adds the necessary support in libgcc to allow using the machine independent 16-bit floating point support. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> libgcc/ @@ -338,7 +329,7 @@ libgcc/ supports _Float16 and __bfloat16 types. * configure: Regenerate. -==================== Branch work247-float, patch #201 ==================== +==================== Branch work247-float, patch #221 ==================== Add initial 16-bit floating point support. @@ -354,7 +345,7 @@ usable, the next patch will also need to be installed. That patch will add support in libgcc for 16-bit floating point support. -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -423,7 +414,7 @@ gcc/ * config/rs6000/rs6000.opt (-mloat16): New option. * doc/invoke.texi (RS/6000 and PowerPC Options): Document -mfloat16. -==================== Branch work247-float, patch #200 ==================== +==================== Branch work247-float, patch #220 ==================== Add infrastructure for _Float16 and __bfloat16 types. @@ -473,7 +464,7 @@ will be added in the next patch. V4HFmode -- 64-bit vector mode _Float16 used in some insns -2026-06-04 Michael Meissner <[email protected]> +2026-06-18 Michael Meissner <[email protected]> gcc/ @@ -556,6 +547,18 @@ gcc/ (vsx_extract_<mode>_store_p9): Add V8HFmode and V8BFmode. (vsx_extract_<mode>_p8): Likewise. +==================== Branch work247-float, patch #210 was reverted ==================== +==================== Branch work247-float, patch #209 was reverted ==================== +==================== Branch work247-float, patch #208 was reverted ==================== +==================== Branch work247-float, patch #207 was reverted ==================== +==================== Branch work247-float, patch #206 was reverted ==================== +==================== Branch work247-float, patch #205 was reverted ==================== +==================== Branch work247-float, patch #204 was reverted ==================== +==================== Branch work247-float, patch #203 was reverted ==================== +==================== Branch work247-float, patch #202 was reverted ==================== +==================== Branch work247-float, patch #201 was reverted ==================== +==================== Branch work247-float, patch #200 was reverted ==================== + ==================== Branch work247-float, baseline ==================== 2026-06-04 Michael Meissner <[email protected]>
