- cmucl-cvs - mailman3.common-lisp.net

[Git][cmucl/cmucl][issue-425-correctly-rounded-math-functions] Revert "Experiment to see if -march=haswell works better on CI."
by Raymond Toy (＠rtoy) 30 Jan '26

30 Jan '26

Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions at cmucl / cmucl Commits: 010f03c6 by Raymond Toy at 2026-01-30T09:25:32-08:00 Revert "Experiment to see if -march=haswell works better on CI." This reverts commit c61e55e6984de4a9c3e380e3b91faf4e86141f34. - - - - - 2 changed files: - src/lisp/Config.x86_linux - src/lisp/Config.x86_linux_clang Changes: ===================================== src/lisp/Config.x86_linux ===================================== @@ -9,8 +9,7 @@ CPPFLAGS += -m32 # Gcc 15.2.1 and possibly other version can't compile core-math # without this option. CFLAGS += -std=gnu11 -#CFLAGS += -rdynamic -march=pentium4 -mfpmath=sse -mtune=generic -CFLAGS += -rdynamic -march=haswell -mfpmath=sse +CFLAGS += -rdynamic -march=pentium4 -mfpmath=sse -mtune=generic # ANALYZER flag should be set to -fanalyzer when doing the CI builds # for the C static analyzer. CFLAGS += -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 $(ANALYZER) ===================================== src/lisp/Config.x86_linux_clang ===================================== @@ -8,8 +8,7 @@ CFLAGS += $(COPT) # Allow sse2 instructions (-msse2); and tune for pentium4 # (-mtune=pentium4), the first chip to have sse2; and finally generate # code assuming instructions can trap (-ftrapping-math). -#CFLAGS += -msse2 -mtune=pentium4 -ftrapping-math -CFLAGS += -msse2 -mtune=haswell -ftrapping-math +CFLAGS += -msse2 -mtune=pentium4 -ftrapping-math CFLAGS += -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 UNDEFSYMPATTERN = -Xlinker -u -Xlinker & View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/010f03c653a1048051089c6… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/010f03c653a1048051089c6… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-425-correctly-rounded-math-functions] Experiment to see if -march=haswell works better on CI.
by Raymond Toy (＠rtoy) 30 Jan '26

30 Jan '26

Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions at cmucl / cmucl Commits: c61e55e6 by Raymond Toy at 2026-01-29T18:31:13-08:00 Experiment to see if -march=haswell works better on CI. - - - - - 2 changed files: - src/lisp/Config.x86_linux - src/lisp/Config.x86_linux_clang Changes: ===================================== src/lisp/Config.x86_linux ===================================== @@ -9,7 +9,8 @@ CPPFLAGS += -m32 # Gcc 15.2.1 and possibly other version can't compile core-math # without this option. CFLAGS += -std=gnu11 -CFLAGS += -rdynamic -march=pentium4 -mfpmath=sse -mtune=generic +#CFLAGS += -rdynamic -march=pentium4 -mfpmath=sse -mtune=generic +CFLAGS += -rdynamic -march=haswell -mfpmath=sse # ANALYZER flag should be set to -fanalyzer when doing the CI builds # for the C static analyzer. CFLAGS += -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 $(ANALYZER) ===================================== src/lisp/Config.x86_linux_clang ===================================== @@ -8,7 +8,8 @@ CFLAGS += $(COPT) # Allow sse2 instructions (-msse2); and tune for pentium4 # (-mtune=pentium4), the first chip to have sse2; and finally generate # code assuming instructions can trap (-ftrapping-math). -CFLAGS += -msse2 -mtune=pentium4 -ftrapping-math +#CFLAGS += -msse2 -mtune=pentium4 -ftrapping-math +CFLAGS += -msse2 -mtune=haswell -ftrapping-math CFLAGS += -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64 UNDEFSYMPATTERN = -Xlinker -u -Xlinker & View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/c61e55e6984de4a9c3e380e… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/c61e55e6984de4a9c3e380e… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl] Pushed new branch issue-468-core-math-signals
by Raymond Toy (＠rtoy) 28 Jan '26

28 Jan '26

Raymond Toy pushed new branch issue-468-core-math-signals at cmucl / cmucl -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/tree/issue-468-core-math-signa… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-425-correctly-rounded-math-functions] Remove the x87 precision stuff from call-out
by Raymond Toy (＠rtoy) 28 Jan '26

28 Jan '26

Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions at cmucl / cmucl Commits: 10c5d82f by Raymond Toy at 2026-01-28T09:19:54-08:00 Remove the x87 precision stuff from call-out It's not needed anymore because of a previous bug where the x87 precision was erroneously set to single. That error has been fixed. All tests pass still. - - - - - 1 changed file: - src/compiler/x86/sse2-c-call.lisp Changes: ===================================== src/compiler/x86/sse2-c-call.lisp ===================================== @@ -36,33 +36,12 @@ :from :eval :to :result) edx) (:temporary (:sc single-stack) temp-single) (:temporary (:sc double-stack) temp-double) - #+core-math - (:temporary (:sc unsigned-stack) save-fpu-cw) - #+core-math - (:temporary (:sc unsigned-stack) fpu-cw) - #+core-math - (:temporary (:sc unsigned-reg :offset esi-offset) temp-cw) (:node-var node) (:vop-var vop) (:save-p t) (:ignore args ecx edx) (:guard (backend-featurep :sse2)) (:generator 0 - #+core-math - (progn - ;; Save the x87 FPU control word. Then modify it to set the - ;; precision bits to 3 for 64-bit mantissas for 80-bit - ;; arithmetic. If we don't some of some special functions - ;; return incorrect values because the x87 precision was set to - ;; single. - (inst fnstcw save-fpu-cw) - (move temp-cw save-fpu-cw) - (inst and temp-cw (dpb 0 (byte 2 8) #xffff)) ; Zap the precision control bits - (inst or temp-cw (dpb 3 (byte 3 8) 0)) ; Set precision control bits - (move fpu-cw temp-cw) - (inst fldcw fpu-cw) ; New CW - ) - (cond ((policy node (> space speed)) (move eax function) (inst call (make-fixup (extern-alien-name "call_into_c") :foreign))) @@ -84,10 +63,7 @@ (inst movss xmm0-tn (ea-for-sf-stack temp-single))) (double-reg (inst fstpd (ea-for-df-stack temp-double)) - (inst movsd xmm0-tn (ea-for-df-stack temp-double))))) - ;; Restore the x87 FPU control settings - #+core-math - (inst fldcw save-fpu-cw))) + (inst movsd xmm0-tn (ea-for-df-stack temp-double))))))) (define-vop (alloc-number-stack-space) (:info amount) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/10c5d82f9f3ac80939604f4… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/10c5d82f9f3ac80939604f4… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-425-correctly-rounded-math-functions] Implement lisp_sincos for fdlibm and core-math
by Raymond Toy (＠rtoy) 28 Jan '26

28 Jan '26

Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions at cmucl / cmucl Commits: f00d0070 by Raymond Toy at 2026-01-28T08:55:25-08:00 Implement lisp_sincos for fdlibm and core-math First, rename lisp_sincos in lisp/sincos.c to cmucl_sincos. Then define lisp_sincos in irrat.c to call cr_sincos or cmucl_sincos depending on whether we're using core-math or not. - - - - - 3 changed files: - src/code/irrat.lisp - src/lisp/irrat.c - src/lisp/sincos.c Changes: ===================================== src/code/irrat.lisp ===================================== @@ -214,9 +214,7 @@ (y1 double-float :out)) (declaim (inline %%sincos)) -(alien:def-alien-routine (#-core-math "lisp_sincos" - #+core-math "cr_sincos" - %%sincos) +(alien:def-alien-routine ("lisp_sincos" %%sincos) c-call:void (x double-float) (s double-float :out) ===================================== src/lisp/irrat.c ===================================== @@ -31,6 +31,7 @@ extern double cr_pow(double, double); extern double cr_hypot(double, double); extern double cr_log1p(double); extern double cr_expm1(double); +extern void cr_sincos(double, double *, double *); #endif @@ -243,3 +244,15 @@ lisp_scalbn(double x, int n) { return fdlibm_scalbn(x, n); } + +void +lisp_sincos (double x, double *s, double *c) +{ +#ifdef FEATURE_CORE_MATH + cr_sincos(x, s, c); +#else + extern void cmucl_sincos(double, double*, double*); + + cmucl_sincos(x, s, c); +#endif +} ===================================== src/lisp/sincos.c ===================================== @@ -13,7 +13,7 @@ * reduction just once. */ void -lisp_sincos (double x, double *s, double *c) +cmucl_sincos (double x, double *s, double *c) { int ix; union { int i[2]; double d; } ux; View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/f00d0070856ee72247eaeed… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/f00d0070856ee72247eaeed… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-425-correctly-rounded-math-functions] 3 commits: Define CORE64_SRCS instead of CORE64_OBJS
by Raymond Toy (＠rtoy) 28 Jan '26

28 Jan '26

Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions at cmucl / cmucl Commits: 6f100c06 by Raymond Toy at 2026-01-28T08:21:07-08:00 Define CORE64_SRCS instead of CORE64_OBJS Define the list of sources for core-math routines. The rules in GNUMakefile handle converting the list of sources into a list of object files. Then we can clean up the makefile a bit to reduce unneeded complexity. - - - - - 08f23a38 by Raymond Toy at 2026-01-28T08:22:47-08:00 Use fdlibm_scalbn even when core-math is used. Core-math doesn't have a scalbn function, but fdlibm does so use that. We could use libm's scalbn, but that doesn't signal errors on overflow like we want. - - - - - 34837462 by Raymond Toy at 2026-01-28T08:28:42-08:00 Disable some tests for core-math This is so we can pass CI. We'll need to come back and look at these again after we decide what we want to do with these failures for core-math which pass when using fdlibm. - - - - - 4 changed files: - src/lisp/Config.x86_common - src/lisp/GNUmakefile - src/lisp/irrat.c - tests/fdlibm.lisp Changes: ===================================== src/lisp/Config.x86_common ===================================== @@ -42,29 +42,29 @@ endif ifdef FEATURE_CORE_MATH CORE_MATH_64=core-math/src/binary64 -CORE64_OBJS=\ - $(CORE_MATH_64)/sin/sin.o \ - $(CORE_MATH_64)/cos/cos.o \ - $(CORE_MATH_64)/tan/tan.o \ - $(CORE_MATH_64)/atan2/atan2.o \ - $(CORE_MATH_64)/asin/asin.o \ - $(CORE_MATH_64)/acos/acos.o \ - $(CORE_MATH_64)/atan/atan.o \ - $(CORE_MATH_64)/sinh/sinh.o \ - $(CORE_MATH_64)/cosh/cosh.o \ - $(CORE_MATH_64)/tanh/tanh.o \ - $(CORE_MATH_64)/asinh/asinh.o \ - $(CORE_MATH_64)/acosh/acosh.o \ - $(CORE_MATH_64)/atanh/atanh.o \ - $(CORE_MATH_64)/exp/exp.o \ - $(CORE_MATH_64)/log/log.o \ - $(CORE_MATH_64)/log10/log10.o \ - $(CORE_MATH_64)/log2/log2.o \ - $(CORE_MATH_64)/pow/pow.o \ - $(CORE_MATH_64)/hypot/hypot.o \ - $(CORE_MATH_64)/log1p/log1p.o \ - $(CORE_MATH_64)/expm1/expm1.o \ - $(CORE_MATH_64)/sincos/sincos.o +CORE64_SRCS=\ + $(CORE_MATH_64)/sin/sin.c \ + $(CORE_MATH_64)/cos/cos.c \ + $(CORE_MATH_64)/tan/tan.c \ + $(CORE_MATH_64)/atan2/atan2.c \ + $(CORE_MATH_64)/asin/asin.c \ + $(CORE_MATH_64)/acos/acos.c \ + $(CORE_MATH_64)/atan/atan.c \ + $(CORE_MATH_64)/sinh/sinh.c \ + $(CORE_MATH_64)/cosh/cosh.c \ + $(CORE_MATH_64)/tanh/tanh.c \ + $(CORE_MATH_64)/asinh/asinh.c \ + $(CORE_MATH_64)/acosh/acosh.c \ + $(CORE_MATH_64)/atanh/atanh.c \ + $(CORE_MATH_64)/exp/exp.c \ + $(CORE_MATH_64)/log/log.c \ + $(CORE_MATH_64)/log10/log10.c \ + $(CORE_MATH_64)/log2/log2.c \ + $(CORE_MATH_64)/pow/pow.c \ + $(CORE_MATH_64)/hypot/hypot.c \ + $(CORE_MATH_64)/log1p/log1p.c \ + $(CORE_MATH_64)/expm1/expm1.c \ + $(CORE_MATH_64)/sincos/sincos.c endif ifeq ($(filter 2% 3%, $(shell $(CC) -dumpversion)),) ===================================== src/lisp/GNUmakefile ===================================== @@ -36,7 +36,7 @@ SRCS = lisp.c coreparse.c alloc.c monitor.c print.c interr.c \ dynbind.c breakpoint.c regnames.c backtrace.c save.c purify.c \ runprog.c time.c case-mapping.c exec-init.c \ irrat.c \ - ${CORE64_OBJS:.o=.c} \ + ${CORE64_SRCS} \ ${FDLIBM} ${ARCH_SRC} ${ASSEM_SRC} ${OS_SRC} ${GC_SRC} @@ -61,8 +61,8 @@ lisp: ${OBJS} ${CORE64_OBJS} version.o # Create a library out of all the object files so we can build an # executable. However, we need to remove exec-init.o from the library -lisp.a: version.o ${OBJS} ${CORE64_OBJS} ${EXEC_FINAL_OBJ} - $(AR) crs lisp.a ${OBJS} ${CORE64_OBJS} version.o +lisp.a: version.o ${OBJS} ${EXEC_FINAL_OBJ} + $(AR) crs lisp.a ${OBJS} version.o ifneq (${EXEC_FINAL_OBJ},) $(AR) d lisp.a exec-init.o endif ===================================== src/lisp/irrat.c ===================================== @@ -7,6 +7,8 @@ #include "lisp.h" #include "internals.h" +#include "fdlibm.h" + #ifdef FEATURE_CORE_MATH #include <math.h> extern double cr_sin(double); @@ -29,8 +31,6 @@ extern double cr_pow(double, double); extern double cr_hypot(double, double); extern double cr_log1p(double); extern double cr_expm1(double); -#else -#include "fdlibm.h" #endif @@ -241,9 +241,5 @@ lisp_expm1(double x) double lisp_scalbn(double x, int n) { -#ifdef FEATURE_CORE_MATH - return scalbn(x, n); -#else return fdlibm_scalbn(x, n); -#endif } ===================================== tests/fdlibm.lisp ===================================== @@ -171,6 +171,7 @@ (ext:with-float-traps-masked (:invalid) (assert-true (ext::float-nan-p (kernel:%expm1 *snan*)))) ;; expm1(x) = -1 for x < -56*log(2), signaling inexact + #-core-math (let ((x (* -57 (log 2d0)))) (ext:with-float-traps-enabled (:inexact) (assert-error 'floating-point-inexact @@ -217,6 +218,7 @@ (ext:with-float-traps-masked (:overflow) (assert-equal ext:double-float-positive-infinity (kernel:%exp 710d0))) + #-core-math (let ((modes (ext:get-floating-point-modes))) (unwind-protect (progn View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/66d306ffceda908c6d56b8… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/66d306ffceda908c6d56b8… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-425-correctly-rounded-math-functions] 6 commits: Fix #466: Add C wrapper for the special functions
by Raymond Toy (＠rtoy) 28 Jan '26

28 Jan '26

Raymond Toy pushed to branch issue-425-correctly-rounded-math-functions at cmucl / cmucl Commits: 7f708431 by Raymond Toy at 2026-01-27T08:45:32-08:00 Fix #466: Add C wrapper for the special functions - - - - - 318d0560 by Raymond Toy at 2026-01-27T08:45:32-08:00 Merge branch 'issue-466-c-wrapper-specfun' into 'master' Fix #466: Add C wrapper for the special functions Closes #466 See merge request cmucl/cmucl!345 - - - - - 8ceef426 by Raymond Toy at 2026-01-27T18:42:27-08:00 Merge branch 'master' into issue-425-correctly-rounded-math-functions - - - - - 5a960d14 by Raymond Toy at 2026-01-27T19:44:00-08:00 Fix up incorrect merge We need to get rid of most of the #+core-math stuff. - - - - - f364498b by Raymond Toy at 2026-01-27T20:01:10-08:00 Add core-math routines to SRCS Just for completeness. - - - - - 66d306ff by Raymond Toy at 2026-01-27T20:02:09-08:00 Call the core-math routines when FEATURE_CORE_MATH is set Otherwise, we use the old fdlibm routines. - - - - - 5 changed files: - src/code/irrat.lisp - src/lisp/GNUmakefile - src/lisp/fdlibm.h - + src/lisp/irrat.c - src/lisp/log2.c Changes: ===================================== src/code/irrat.lisp ===================================== @@ -75,73 +75,38 @@ ;;; Please refer to the Unix man pages for details about these routines. -#+core-math -(progn -;;; Trigonometric. -(def-math-rtn ("cr_sin" %sin) 1) -(def-math-rtn ("cr_cos" %cos) 1) -(def-math-rtn ("cr_tan" %tan) 1) -(def-math-rtn ("cr_atan" %atan) 1) -(def-math-rtn ("cr_atan2" %atan2) 2) -(def-math-rtn ("cr_asin" %asin) 1) -(def-math-rtn ("cr_acos" %acos) 1) -(def-math-rtn ("cr_sinh" %sinh) 1) -(def-math-rtn ("cr_cosh" %cosh) 1) -(def-math-rtn ("cr_tanh" %tanh) 1) -(def-math-rtn ("cr_asinh" %asinh) 1) -(def-math-rtn ("cr_acosh" %acosh) 1) -(def-math-rtn ("cr_atanh" %atanh) 1) - -;;; Exponential and Logarithmic. -(def-math-rtn ("cr_exp" %exp) 1) -(def-math-rtn ("cr_log" %log) 1) -(def-math-rtn ("cr_log10" %log10) 1) -(def-math-rtn ("cr_log2" %log2) 1) - -(def-math-rtn ("__ieee754_pow" %pow) 2) -#-(or x86 sparc-v7 sparc-v8 sparc-v9) -(def-math-rtn "sqrt" 1) -(def-math-rtn ("cr_hypot" %hypot) 2) - -(def-math-rtn ("cr_log1p" %log1p) 1) -(def-math-rtn ("cr_expm1" %expm1) 1) -) - -#-core-math -(progn ;;; Trigonometric. -(def-math-rtn ("fdlibm_sin" %sin) 1) -(def-math-rtn ("fdlibm_cos" %cos) 1) -(def-math-rtn ("fdlibm_tan" %tan) 1) -(def-math-rtn ("fdlibm_atan" %atan) 1) -(def-math-rtn ("__ieee754_atan2" %atan2) 2) -(def-math-rtn ("__ieee754_asin" %asin) 1) -(def-math-rtn ("__ieee754_acos" %acos) 1) -(def-math-rtn ("__ieee754_sinh" %sinh) 1) -(def-math-rtn ("__ieee754_cosh" %cosh) 1) -(def-math-rtn ("fdlibm_tanh" %tanh) 1) -(def-math-rtn ("fdlibm_asinh" %asinh) 1) -(def-math-rtn ("__ieee754_acosh" %acosh) 1) -(def-math-rtn ("__ieee754_atanh" %atanh) 1) +(def-math-rtn ("lisp_sin" %sin) 1) +(def-math-rtn ("lisp_cos" %cos) 1) +(def-math-rtn ("lisp_tan" %tan) 1) +(def-math-rtn ("lisp_atan" %atan) 1) +(def-math-rtn ("lisp_atan2" %atan2) 2) +(def-math-rtn ("lisp_asin" %asin) 1) +(def-math-rtn ("lisp_acos" %acos) 1) +(def-math-rtn ("lisp_sinh" %sinh) 1) +(def-math-rtn ("lisp_cosh" %cosh) 1) +(def-math-rtn ("lisp_tanh" %tanh) 1) +(def-math-rtn ("lisp_asinh" %asinh) 1) +(def-math-rtn ("lisp_acosh" %acosh) 1) +(def-math-rtn ("lisp_atanh" %atanh) 1) ;;; Exponential and Logarithmic. -(def-math-rtn ("__ieee754_exp" %exp) 1) -(def-math-rtn ("__ieee754_log" %log) 1) -(def-math-rtn ("__ieee754_log10" %log10) 1) -(def-math-rtn ("cmucl_log2" %log2) 1) +(def-math-rtn ("lisp_exp" %exp) 1) +(def-math-rtn ("lisp_log" %log) 1) +(def-math-rtn ("lisp_log10" %log10) 1) +(def-math-rtn ("lisp_log2" %log2) 1) -(def-math-rtn ("__ieee754_pow" %pow) 2) +(def-math-rtn ("lisp_pow" %pow) 2) #-(or x86 sparc-v7 sparc-v8 sparc-v9) (def-math-rtn "sqrt" 1) -(def-math-rtn ("__ieee754_hypot" %hypot) 2) +(def-math-rtn ("lisp_hypot" %hypot) 2) -(def-math-rtn ("fdlibm_log1p" %log1p) 1) -(def-math-rtn ("fdlibm_expm1" %expm1) 1) -) +(def-math-rtn ("lisp_log1p" %log1p) 1) +(def-math-rtn ("lisp_expm1" %expm1) 1) (declaim (inline %scalbn)) (export '%scalbn) -(alien:def-alien-routine ("fdlibm_scalbn" %scalbn) c-call:double +(alien:def-alien-routine ("lisp_scalbn" %scalbn) c-call:double (x double-float) (n c-call:int)) ===================================== src/lisp/GNUmakefile ===================================== @@ -35,6 +35,8 @@ SRCS = lisp.c coreparse.c alloc.c monitor.c print.c interr.c \ vars.c parse.c interrupt.c search.c validate.c globals.c \ dynbind.c breakpoint.c regnames.c backtrace.c save.c purify.c \ runprog.c time.c case-mapping.c exec-init.c \ + irrat.c \ + ${CORE64_OBJS:.o=.c} \ ${FDLIBM} ${ARCH_SRC} ${ASSEM_SRC} ${OS_SRC} ${GC_SRC} ===================================== src/lisp/fdlibm.h ===================================== @@ -54,8 +54,21 @@ extern double fdlibm_tan(double x); extern double fdlibm_expm1(double x); extern double fdlibm_log1p(double x); extern double fdlibm_atan(double x); +extern double fdlibm_tanh(double x); +extern double fdlibm_asinh(double x); extern double __ieee754_exp(double x); extern double __ieee754_log(double x); +extern double __ieee754_atan2(double y, double x); +extern double __ieee754_asin(double x); +extern double __ieee754_acos(double x); +extern double __ieee754_sinh(double x); +extern double __ieee754_cosh(double x); +extern double __ieee754_atanh(double x); +extern double __ieee754_acosh(double x); +extern double __ieee754_log10(double x); +extern double __ieee754_pow(double x, double y); +extern double __ieee754_hypot(double x, double y); +extern double fdlibm_scalbn(double x, int n); enum FDLIBM_EXCEPTION { FDLIBM_DIVIDE_BY_ZERO, ===================================== src/lisp/irrat.c ===================================== @@ -0,0 +1,249 @@ +/* + + This code was written as part of the CMU Common Lisp project at + Carnegie Mellon University, and has been placed in the public domain. + +*/ + +#include "lisp.h" +#include "internals.h" +#ifdef FEATURE_CORE_MATH +#include <math.h> +extern double cr_sin(double); +extern double cr_cos(double); +extern double cr_tan(double); +extern double cr_atan(double); +extern double cr_atan2(double, double); +extern double cr_asin(double); +extern double cr_acos(double); +extern double cr_sinh(double); +extern double cr_cosh(double); +extern double cr_tanh(double); +extern double cr_asinh(double); +extern double cr_acosh(double); +extern double cr_atanh(double); +extern double cr_exp(double); +extern double cr_log(double); +extern double cr_log10(double); +extern double cr_pow(double, double); +extern double cr_hypot(double, double); +extern double cr_log1p(double); +extern double cr_expm1(double); +#else +#include "fdlibm.h" +#endif + + +/* + * Wrappers for the special functions + */ + +double +lisp_sin(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_sin(x); +#else + return fdlibm_sin(x); +#endif +} + +double +lisp_cos(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_cos(x); +#else + return fdlibm_cos(x); +#endif +} + +double +lisp_tan(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_tan(x); +#else + return fdlibm_tan(x); +#endif +} + +double +lisp_atan(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_atan(x); +#else + return fdlibm_atan(x); +#endif +} + +double +lisp_atan2(double y, double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_atan2(y, x); +#else + return __ieee754_atan2(y, x); +#endif +} + +double +lisp_asin(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_asin(x); +#else + return __ieee754_asin(x); +#endif +} + +double +lisp_acos(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_acos(x); +#else + return __ieee754_acos(x); +#endif +} + +double +lisp_sinh(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_sinh(x); +#else + return __ieee754_sinh(x); +#endif +} + +double +lisp_cosh(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_cosh(x); +#else + return __ieee754_cosh(x); +#endif +} + +double +lisp_tanh(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_tanh(x); +#else + return fdlibm_tanh(x); +#endif +} + +double +lisp_asinh(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_asinh(x); +#else + return fdlibm_asinh(x); +#endif +} + +double +lisp_acosh(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_acosh(x); +#else + return __ieee754_acosh(x); +#endif +} + +double +lisp_atanh(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_atanh(x); +#else + return __ieee754_atanh(x); +#endif +} + +double +lisp_exp(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_exp(x); +#else + return __ieee754_exp(x); +#endif +} + +double +lisp_log(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_log(x); +#else + return __ieee754_log(x); +#endif +} + +double +lisp_log10(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_log10(x); +#else + return __ieee754_log10(x); +#endif +} + +double +lisp_pow(double x, double y) +{ +#ifdef FEATURE_CORE_MATH + return cr_pow(x, y); +#else + return __ieee754_pow(x, y); +#endif +} + +double +lisp_hypot(double x, double y) +{ +#ifdef FEATURE_CORE_MATH + return cr_hypot(x, y); +#else + return __ieee754_hypot(x, y); +#endif +} + +double +lisp_log1p(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_log1p(x); +#else + return fdlibm_log1p(x); +#endif +} + +double +lisp_expm1(double x) +{ +#ifdef FEATURE_CORE_MATH + return cr_expm1(x); +#else + return fdlibm_expm1(x); +#endif +} + +double +lisp_scalbn(double x, int n) +{ +#ifdef FEATURE_CORE_MATH + return scalbn(x, n); +#else + return fdlibm_scalbn(x, n); +#endif +} ===================================== src/lisp/log2.c ===================================== @@ -39,7 +39,7 @@ cp = 9.61796693925975554329e-01, /* 0x3FEEC709, 0xDC3A03FD =2/(3ln2) */ cp_h = 9.61796700954437255859e-01, /* 0x3FEEC709, 0xE0000000 =(float)cp */ cp_l = -7.02846165095275826516e-09; /* 0xBE3E2FE0, 0x145B01F5 =tail of cp_h*/ -double cmucl_log2(double x) +double lisp_log2(double x) { double ax; int k, hx, lx, ix; View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/a5440432136bbcc0de718c… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/a5440432136bbcc0de718c… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][master] 2 commits: Fix #466: Add C wrapper for the special functions
by Raymond Toy (＠rtoy) 27 Jan '26

27 Jan '26

Raymond Toy pushed to branch master at cmucl / cmucl Commits: 7f708431 by Raymond Toy at 2026-01-27T08:45:32-08:00 Fix #466: Add C wrapper for the special functions - - - - - 318d0560 by Raymond Toy at 2026-01-27T08:45:32-08:00 Merge branch 'issue-466-c-wrapper-specfun' into 'master' Fix #466: Add C wrapper for the special functions Closes #466 See merge request cmucl/cmucl!345 - - - - - 5 changed files: - src/code/irrat.lisp - src/lisp/GNUmakefile - src/lisp/fdlibm.h - + src/lisp/irrat.c - src/lisp/log2.c Changes: ===================================== src/code/irrat.lisp ===================================== @@ -76,37 +76,37 @@ ;;; Please refer to the Unix man pages for details about these routines. ;;; Trigonometric. -(def-math-rtn ("fdlibm_sin" %sin) 1) -(def-math-rtn ("fdlibm_cos" %cos) 1) -(def-math-rtn ("fdlibm_tan" %tan) 1) -(def-math-rtn ("fdlibm_atan" %atan) 1) -(def-math-rtn ("__ieee754_atan2" %atan2) 2) -(def-math-rtn ("__ieee754_asin" %asin) 1) -(def-math-rtn ("__ieee754_acos" %acos) 1) -(def-math-rtn ("__ieee754_sinh" %sinh) 1) -(def-math-rtn ("__ieee754_cosh" %cosh) 1) -(def-math-rtn ("fdlibm_tanh" %tanh) 1) -(def-math-rtn ("fdlibm_asinh" %asinh) 1) -(def-math-rtn ("__ieee754_acosh" %acosh) 1) -(def-math-rtn ("__ieee754_atanh" %atanh) 1) +(def-math-rtn ("lisp_sin" %sin) 1) +(def-math-rtn ("lisp_cos" %cos) 1) +(def-math-rtn ("lisp_tan" %tan) 1) +(def-math-rtn ("lisp_atan" %atan) 1) +(def-math-rtn ("lisp_atan2" %atan2) 2) +(def-math-rtn ("lisp_asin" %asin) 1) +(def-math-rtn ("lisp_acos" %acos) 1) +(def-math-rtn ("lisp_sinh" %sinh) 1) +(def-math-rtn ("lisp_cosh" %cosh) 1) +(def-math-rtn ("lisp_tanh" %tanh) 1) +(def-math-rtn ("lisp_asinh" %asinh) 1) +(def-math-rtn ("lisp_acosh" %acosh) 1) +(def-math-rtn ("lisp_atanh" %atanh) 1) ;;; Exponential and Logarithmic. -(def-math-rtn ("__ieee754_exp" %exp) 1) -(def-math-rtn ("__ieee754_log" %log) 1) -(def-math-rtn ("__ieee754_log10" %log10) 1) -(def-math-rtn ("cmucl_log2" %log2) 1) +(def-math-rtn ("lisp_exp" %exp) 1) +(def-math-rtn ("lisp_log" %log) 1) +(def-math-rtn ("lisp_log10" %log10) 1) +(def-math-rtn ("lisp_log2" %log2) 1) -(def-math-rtn ("__ieee754_pow" %pow) 2) +(def-math-rtn ("lisp_pow" %pow) 2) #-(or x86 sparc-v7 sparc-v8 sparc-v9) (def-math-rtn "sqrt" 1) -(def-math-rtn ("__ieee754_hypot" %hypot) 2) +(def-math-rtn ("lisp_hypot" %hypot) 2) -(def-math-rtn ("fdlibm_log1p" %log1p) 1) -(def-math-rtn ("fdlibm_expm1" %expm1) 1) +(def-math-rtn ("lisp_log1p" %log1p) 1) +(def-math-rtn ("lisp_expm1" %expm1) 1) (declaim (inline %scalbn)) (export '%scalbn) -(alien:def-alien-routine ("fdlibm_scalbn" %scalbn) c-call:double +(alien:def-alien-routine ("lisp_scalbn" %scalbn) c-call:double (x double-float) (n c-call:int)) ===================================== src/lisp/GNUmakefile ===================================== @@ -35,6 +35,7 @@ SRCS = lisp.c coreparse.c alloc.c monitor.c print.c interr.c \ vars.c parse.c interrupt.c search.c validate.c globals.c \ dynbind.c breakpoint.c regnames.c backtrace.c save.c purify.c \ runprog.c time.c case-mapping.c exec-init.c \ + irrat.c \ ${FDLIBM} ${ARCH_SRC} ${ASSEM_SRC} ${OS_SRC} ${GC_SRC} OBJS = $(patsubst %.c,%.o,$(patsubst %.S,%.o,$(patsubst %.s,%.o,$(SRCS)))) ===================================== src/lisp/fdlibm.h ===================================== @@ -54,8 +54,21 @@ extern double fdlibm_tan(double x); extern double fdlibm_expm1(double x); extern double fdlibm_log1p(double x); extern double fdlibm_atan(double x); +extern double fdlibm_tanh(double x); +extern double fdlibm_asinh(double x); extern double __ieee754_exp(double x); extern double __ieee754_log(double x); +extern double __ieee754_atan2(double y, double x); +extern double __ieee754_asin(double x); +extern double __ieee754_acos(double x); +extern double __ieee754_sinh(double x); +extern double __ieee754_cosh(double x); +extern double __ieee754_atanh(double x); +extern double __ieee754_acosh(double x); +extern double __ieee754_log10(double x); +extern double __ieee754_pow(double x, double y); +extern double __ieee754_hypot(double x, double y); +extern double fdlibm_scalbn(double x, int n); enum FDLIBM_EXCEPTION { FDLIBM_DIVIDE_BY_ZERO, ===================================== src/lisp/irrat.c ===================================== @@ -0,0 +1,138 @@ +/* + + This code was written as part of the CMU Common Lisp project at + Carnegie Mellon University, and has been placed in the public domain. + +*/ + +#include "fdlibm.h" + +/* + * Wrappers for the special functions + */ + +double +lisp_sin(double x) +{ + return fdlibm_sin(x); +} + +double +lisp_cos(double x) +{ + return fdlibm_cos(x); +} + +double +lisp_tan(double x) +{ + return fdlibm_tan(x); +} + +double +lisp_atan(double x) +{ + return fdlibm_atan(x); +} + +double +lisp_atan2(double y, double x) +{ + return __ieee754_atan2(y, x); +} + +double +lisp_asin(double x) +{ + return __ieee754_asin(x); +} + +double +lisp_acos(double x) +{ + return __ieee754_acos(x); +} + +double +lisp_sinh(double x) +{ + return __ieee754_sinh(x); +} + +double +lisp_cosh(double x) +{ + return __ieee754_cosh(x); +} + +double +lisp_tanh(double x) +{ + return fdlibm_tanh(x); +} + +double +lisp_asinh(double x) +{ + return fdlibm_asinh(x); +} + +double +lisp_acosh(double x) +{ + return __ieee754_acosh(x); +} + +double +lisp_atanh(double x) +{ + return __ieee754_atanh(x); +} + +double +lisp_exp(double x) +{ + return __ieee754_exp(x); +} + +double +lisp_log(double x) +{ + return __ieee754_log(x); +} + +double +lisp_log10(double x) +{ + return __ieee754_log10(x); +} + +double +lisp_pow(double x, double y) +{ + return __ieee754_pow(x, y); +} + +double +lisp_hypot(double x, double y) +{ + return __ieee754_hypot(x, y); +} + +double +lisp_log1p(double x) +{ + return fdlibm_log1p(x); +} + +double +lisp_expm1(double x) +{ + return fdlibm_expm1(x); +} + +double +lisp_scalbn(double x, int n) +{ + return fdlibm_scalbn(x, n); +} ===================================== src/lisp/log2.c ===================================== @@ -39,7 +39,7 @@ cp = 9.61796693925975554329e-01, /* 0x3FEEC709, 0xDC3A03FD =2/(3ln2) */ cp_h = 9.61796700954437255859e-01, /* 0x3FEEC709, 0xE0000000 =(float)cp */ cp_l = -7.02846165095275826516e-09; /* 0xBE3E2FE0, 0x145B01F5 =tail of cp_h*/ -double cmucl_log2(double x) +double lisp_log2(double x) { double ax; int k, hx, lx, ix; View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/ea927341913119e29bf891… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/compare/ea927341913119e29bf891… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-466-c-wrapper-specfun] Fix typo
by Raymond Toy (＠rtoy) 27 Jan '26

27 Jan '26

Raymond Toy pushed to branch issue-466-c-wrapper-specfun at cmucl / cmucl Commits: e2cb39cc by Raymond Toy at 2026-01-27T07:47:05-08:00 Fix typo - - - - - 1 changed file: - src/code/irrat.lisp Changes: ===================================== src/code/irrat.lisp ===================================== @@ -49,7 +49,7 @@ (string-upcase name))))) `(progn (declaim (inline ,lisp-name)) - (export ,lisp-name) + (export ',lisp-name) (alien:def-alien-routine (,c-name ,lisp-name) double-float ,@(let ((results nil)) (dotimes (i num-args (nreverse results)) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/e2cb39cc083c3156b760522… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/e2cb39cc083c3156b760522… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0

[Git][cmucl/cmucl][issue-466-c-wrapper-specfun] Finish reversion of single-float changes
by Raymond Toy (＠rtoy) 27 Jan '26

27 Jan '26

Raymond Toy pushed to branch issue-466-c-wrapper-specfun at cmucl / cmucl Commits: c55685ee by Raymond Toy at 2026-01-27T07:42:55-08:00 Finish reversion of single-float changes Remove the parens in the export in def-math-rtn. This returns us to what was there originally instead of being gratuitously different. - - - - - 1 changed file: - src/code/irrat.lisp Changes: ===================================== src/code/irrat.lisp ===================================== @@ -49,7 +49,7 @@ (string-upcase name))))) `(progn (declaim (inline ,lisp-name)) - (export '(,lisp-name)) + (export ,lisp-name) (alien:def-alien-routine (,c-name ,lisp-name) double-float ,@(let ((results nil)) (dotimes (i num-args (nreverse results)) View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/c55685eeac58664c350a026… -- View it on GitLab: https://gitlab.common-lisp.net/cmucl/cmucl/-/commit/c55685eeac58664c350a026… You're receiving this email because of your account on gitlab.common-lisp.net.

1 0