/[cln]/ChangeLog
ViewVC logotype

Diff of /ChangeLog

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 1.138 by kreckel, Tue Aug 30 21:32:52 2005 UTC revision 1.151 by haible, Wed Apr 19 16:10:18 2006 UTC
# Line 1  Line 1 
1    2006-04-19  Bruno Haible  <bruno@clisp.org>
2    
3            Prepare for autoconf-2.60.
4            * general.m4 (CL_CC_WORKS): Include <stdlib.h>, for exit() declaration.
5            * longdouble.m4 (CL_LONGDOUBLE): Likewise.
6            * longlong.m4 (CL_LONGLONG): Likewise.
7            * times.m4 (CL_TIMES_CLOCK): Likewise.
8            Reported by Ralf Wildenhues <Ralf.Wildenhues@gmx.de>.
9    
10    2005-12-04  Bruno Haible  <bruno@clisp.org>
11    
12            * src/integer/conv/cl_I_to_digits.cc (I_to_digits_noshrink): Set
13            erg->len at the end.
14    
15    2005-12-04  Bruno Haible  <bruno@clisp.org>
16    
17            Extend the fixnum range from 32 bits to 61 bits on 64-bit platforms.
18            * doc/cln.tex (gcd, jacobi): Take 'long' instead of 'int32' arguments.
19            * include/cln/object.h (cl_value_shift): Define as 3, not 32, in the
20            64-bit case.
21            * include/cln/types.h (intVsize): New macro.
22            (sintV, uintV): New types.
23            * include/cln/integer.h (gcd): Take uintV arguments.
24            * include/cln/numtheory.h (jacobi): Take sintV arguments.
25            * src/complex/input/cl_N_read.cc (read_complex): Call FN_to_UV instead
26            of FN_to_UL.
27            * src/complex/transcendental/cl_C_expt_C.cc: Likewise.
28            * src/float/dfloat/elem/cl_DF_scale_I.cc: Use uintV instead of uintL
29            * src/float/dfloat/elem/cl_DF_from_RA.cc: Don't assume values >= 2^53
30            and < 2^55 are always bignums.
31            where appropriate.
32            * src/float/ffloat/conv/cl_RA_to_float.cc: Call FN_to_UV instead of
33            FN_to_UL.
34            * src/float/ffloat/elem/cl_FF_from_RA.cc: Likewise.
35            * src/float/ffloat/elem/cl_FF_scale_I.cc: Call FN_to_V instead of
36            FN_to_L.
37            * src/float/lfloat/elem/cl_LF_scale_I.cc: Likewise.
38            * src/float/output/cl_F_dprint.cc: Likewise.
39            * src/float/sfloat/elem/cl_SF_from_RA.cc: Call FN_to_UV instead of
40            FN_to_UL.
41            * src/float/sfloat/elem/cl_SF_scale_I.cc: Call FN_to_V instead of
42            FN_to_L.
43            * src/integer/cl_I.h (FN_to_UV): Renamed from FN_to_UL, change return
44            type.
45            (FN_to_V): Renamed from FN_to_L, change return type.
46            (FN_V_zerop): Renamed from FN_L_zerop.
47            (FN_V_minusp): Renamed from FN_L_minusp.
48            (cl_I_constructor_from_L2, cl_I_constructor_from_UL2): Define as an
49            inline function on 64-bit platforms.
50            (V_to_I, UV_to_I): New macros.
51            (pFN_maxlength_digits_at): Return an uintV instead of an uint32.
52            (set_pFN_maxlength_digits_at): Take an uintV instead of an uint32.
53            * src/integer/algebraic/cl_I_sqrtp.cc: Call FN_to_UV instead of
54            FN_to_UL.
55            * src/integer/bitwise/cl_I_ash_I.cc: Call FN_to_V instead of FN_to_L.
56            * src/integer/bitwise/cl_I_ilength.cc: Likewise.
57            * src/integer/bitwise/cl_I_log_aux.cc: Likewise.
58            * src/integer/bitwise/cl_I_logbitp_I.cc: Likewise.
59            * src/integer/bitwise/cl_I_logcount.cc: Likewise.
60            * src/integer/bitwise/cl_I_logtest.cc: Likewise.
61            * src/integer/conv/cl_I_from_L2.cc: Don't produce code on 64-bit
62            platforms.
63            * src/integer/conv/cl_I_from_UL2.cc: Likewise.
64            * src/integer/conv/cl_I_from_NDS.cc: Update.
65            * src/integer/conv/cl_I_from_Q.cc: Remove cast to sint32.
66            * src/integer/conv/cl_I_from_UQ.cc: Remove cast to uint32.
67            * src/integer/conv/cl_I_to_L.cc: Check again fixnum that needs more
68            than 32 bits.
69            * src/integer/conv/cl_I_to_Q.cc: Call FN_to_V instead of FN_to_L.
70            * src/integer/conv/cl_I_to_UL.cc: Likewise.
71            * src/integer/conv/cl_I_to_UQ.cc: Likewise.
72            * src/integer/elem/cl_I_div.cc: Treat fixnums that need more than 32
73            bits like bignums.
74            * src/integer/elem/cl_I_minus.cc: Call FN_to_V instead of FN_to_L.
75            * src/integer/elem/cl_I_mul.cc: Likewise.
76            * src/integer/elem/cl_I_plus.cc: Likewise.
77            * src/integer/elem/cl_I_square.cc: Likewise.
78            * src/integer/elem/cl_I_uminus.cc: Likewise.
79            * src/integer/gcd/cl_I_gcd.cc: Likewise.
80            * src/integer/gcd/cl_low_gcd.cc (gcd): Take uintV arguments.
81            * src/integer/hash/cl_I_hashcode.cc: Call FN_to_V instead of FN_to_L.
82            * src/integer/input/cl_I_read.cc (read_integer): Call FN_to_UV instead
83            of FN_to_UL.
84            * src/integer/misc/cl_I_ord2.cc (ord2): Call ord2_64 on 64-bit
85            platforms.
86            * src/integer/misc/cl_I_power2p.cc: Call FN_to_UV instead of FN_to_UL.
87            * src/integer/misc/combin/cl_I_doublefactorial.cc (doublefakul_table):
88            Extend table for larger fixnums.
89            (doublefactorial): Update.
90            * src/integer/misc/combin/cl_I_factorial.cc (fakul_table): Extend table
91            for larger fixnums.
92            (factorial): Update.
93            * src/modinteger/cl_MI_fix16.h: Call FN_to_UV instead of FN_to_UL.
94            * src/modinteger/cl_MI_fix29.h: Likewise.
95            * src/modinteger/cl_MI_fix32.h: Likewise.
96            * src/modinteger/cl_MI_std.h: Likewise.
97            * src/numtheory/cl_nt_cornacchia4.cc: Call FN_to_V instead of FN_to_L.
98            * src/numtheory/cl_nt_jacobi.cc: Likewise.
99            * src/numtheory/cl_nt_jacobi_low.cc (jacobi_aux): Take uintV arguments
100            instead of uint32 arguments.
101            (jacobi): Take sintV argument instead of a sint32 argument.
102            * src/rational/input/cl_RA_read.cc: Call FN_to_UV instead of FN_to_UL.
103            * src/real/input/cl_R_read.cc: Likewise.
104            * src/vector/cl_GV_I.cc: Likewise.
105            * tests/timefact.cc: Call FN_to_V instead of FN_to_L.
106    
107    2005-12-04  Bruno Haible  <bruno@clisp.org>
108    
109            More complete 64-bit division macros.
110            * src/base/cl_low.h (divu_6432_3232_w): Choose a different macro
111            expansion on x86_64.
112            (divu_6432_6432): New macro.
113            (divu_6464_6464): Choose a different macro expansion for all CPUs
114            except sparc64 and x86_64.
115            (divu_12864_6464): Define NEED_VAR_divu_64_rest here.
116            * src/base/low/cl_low_div.cc (divu_64_rest): Avoid defining it twice.
117            (divu_6464_6464_): New function.
118    
119    2005-12-04  Bruno Haible  <bruno@clisp.org>
120    
121            * src/base/cl_low.h (ord2_64): New macro.
122    
123    2005-12-02  Bruno Haible  <bruno@clisp.org>
124    
125            * src/base/cl_low.h (mulu64) [x86_64]: Change asm restriction, since
126            mulq doesn't accept immediate arguments.
127    
128    2005-11-26  Bruno Haible  <bruno@clisp.org>
129    
130            * src/base/cl_low.h (GENERIC_INTEGERLENGTH32): New macro.
131            (integerlength64): Define using integerlength32 if integerlength32
132            is not defined generically.
133    
134    2005-11-26  Bruno Haible  <bruno@clisp.org>
135    
136            * src/base/cl_low.h (mulu32) [SPARC64]: Remove rd instruction, since
137            umul returns the complete 64-bit product in a register.
138            (mulu32_w) [SPARC64]: Prefer umul over mulx instruction.
139            (divu_6432_3232_w) [SPARC64]: Prefer umul/udiv over mulx/udivx
140            instructions.
141    
142    2005-11-26  Bruno Haible  <bruno@clisp.org>
143    
144            * src/base/cl_low.h (divu_3216_1616): Prepend underscore to local
145            variables.
146    
147    2005-11-26  Bruno Haible  <bruno@clisp.org>
148    
149            * src/base/cl_low.h (ord2_32): Parenthesize macro argument.
150    
151    2005-12-17  Richard B. Kreckel  <kreckel@ginac.de>
152    
153            * Created branch cln_1-1 for maintenance patches.
154            This is the main branch, which will eventually become CLN 1.2.0.
155    
156    2005-12-15  Dmitry V. Kustov  <kustov@telex221.ru>
157    
158            * src/base/random/cl_random_from.cc: Add support for OpenBSD.
159    
160    2005-11-23  Richard B. Kreckel  <kreckel@ginac.de>
161    
162            * Version 1.1.11 released.
163    
164    2005-11-20  Richard B. Kreckel  <kreckel@ginac.de>
165    
166            * src/integer/conv/cl_I_cached_power.h: New file.
167            * src/integer/conv/cl_I_cached_power.cc: New file.
168            Contains power_table and cached_power_table previously...
169            * src/integer/conv/cl_I_to_digits.cc: ...here.
170            * src/integer/conv/cl_I_from_digits.cc: Use cached powers.
171    
172    2005-11-02  Richard B. Kreckel  <kreckel@ginac.de>
173    
174            * src/integer/conv/cl_I_from_digits.cc: Made input of all numbers in
175            non-power-of-two base much faster.
176            * tests/test_I_io.cc: New file...
177            * tests/Makefile.in, tests/test_I.cc: ...used here.
178    
179    2005-10-22  Richard B. Kreckel  <kreckel@ginac.de>
180    
181            * Version 1.1.10 released.
182    
183    2005-10-22  Richard B. Kreckel  <kreckel@ginac.de>
184    
185            * src/Makefile.in: Accept CPPFLAGS from environment.
186            * examples/Makefile.in: Likewise.
187            * benchmarks/Makefile.in: Likewise.
188            * tests/Makefile.in: Likewise.
189    
190    2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>
191    
192            * include/cln/modules.h (CL_OUTPUT_LABEL): Work around redundant
193            duplication of basic blocks on m68k.
194    
195  2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>  2005-08-30  Richard B. Kreckel  <kreckel@ginac.de>
196    
197          * include/cln/modules.h (CL_JUMP_TO): Fix mips* brokenness.          * include/cln/modules.h (CL_JUMP_TO): Fix mips* brokenness.
198    
199  2005-08-27  Bruno Haible  <bruno@clisp.org>  2005-08-27  Bruno Haible  <bruno@clisp.org>
200    
201            Split aclocal.m4 into individual files.
202            * m4/alloca.m4, m4/as-underscore.m4, m4/c++-constructors.m4:
203            * m4/fpu_control.m4, m4/general.m4, m4/gettimeofday.m4:
204            * m4/longdouble.m4, m4/longlong.m4, m4/param.m4, m4/perror.m4:
205            * m4/proto.m4, m4/rusage.m4, m4/times.m4:
206            New files, extracted from autoconf/aclocal.m4.
207            * autoconf/aclocal.m4: m4_include them.
208            * Makefile.devel (AUTOCONF_MACROS): New variable.
209            (configure): Depend on it.
210            (CLISP_M4DIR): Remove variable.
211            (autoconf/aclocal.m4): Remove rule.
212    
213    2005-08-27  Bruno Haible  <bruno@clisp.org>
214    
215          * src/integer/bitwise/cl_I_ash_I.cc (ash): Avoid shifting a 32-bit          * src/integer/bitwise/cl_I_ash_I.cc (ash): Avoid shifting a 32-bit
216          zero value by more than 31 bits.          zero value by more than 31 bits.
217    
# Line 15  Line 223 
223    
224  2005-08-27  Bruno Haible  <bruno@clisp.org>  2005-08-27  Bruno Haible  <bruno@clisp.org>
225    
226          * modules.h (CL_OUTPUT_LABEL): Work around redundant duplication          * include/cln/modules.h (CL_OUTPUT_LABEL): Work around redundant
227          of basic blocks by g++ 4.0.          duplication of basic blocks by g++ 4.0.
228          See <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23345>.          See <http://gcc.gnu.org/bugzilla/show_bug.cgi?id=23345>.
229    
230  2005-08-27  Bruno Haible  <bruno@clisp.org>  2005-08-27  Bruno Haible  <bruno@clisp.org>
Legend:
Removed from v.1.138  
changed lines
  Added in v.1.151
Christian Bauer">Christian Bauer
 ViewVC Help
Powered by ViewVC 1.1.15