Merge pull request #450 from libtom/sqr-opt

make mp_sqr and mp_div_3 internal and add optimizations

Author: sjaeckel

demo/test.c

@@ -1241,14 +1241,14 @@ static int test_mp_reduce_2k(void)
    return EXIT_FAILURE;
 }
 
-static int test_mp_div_3(void)
+static int test_s_mp_div_3(void)
 {
    int cnt;
 
    mp_int a, b, c, d, e;
    DOR(mp_init_multi(&a, &b, &c, &d, &e, NULL));
 
-   /* test mp_div_3  */
+   /* test s_mp_div_3  */
    mp_set(&d, 3u);
    for (cnt = 0; cnt < 10000;) {
       mp_digit r2;
@@ -1259,10 +1259,10 @@ static int test_mp_div_3(void)
       }
       DO(mp_rand(&a, (abs(rand_int()) % 128) + 1));
       DO(mp_div(&a, &d, &b, &e));
-      DO(mp_div_3(&a, &c, &r2));
+      DO(s_mp_div_3(&a, &c, &r2));
 
       if (mp_cmp(&b, &c) || mp_cmp_d(&e, r2)) {
-         printf("\nmp_div_3 => Failure\n");
+         printf("\ns_mp_div_3 => Failure\n");
          goto LBL_ERR;
       }
    }
@@ -2297,7 +2297,7 @@ static int unit_tests(int argc, char **argv)
       T1(mp_cnt_lsb, MP_CNT_LSB),
       T1(mp_complement, MP_COMPLEMENT),
       T1(mp_decr, MP_SUB_D),
-      T1(mp_div_3, MP_DIV_3),
+      T1(s_mp_div_3, S_MP_DIV_3),
       T1(mp_dr_reduce, MP_DR_REDUCE),
       T2(mp_pack_unpack,MP_PACK, MP_UNPACK),
       T2(mp_fread_fwrite, MP_FREAD, MP_FWRITE),

doc/bn.tex

@@ -2605,14 +2605,6 @@ \section{Single Digit Functions}
 mp_err mp_decr(mp_int *a);
 \end{alltt}
 
-The division by three can be made faster by replacing the division with a multiplication by the
-multiplicative inverse of three.
-
-\index{mp\_div\_3}
-\begin{alltt}
-mp_err mp_div_3(const mp_int *a, mp_int *c, mp_digit *d);
-\end{alltt}
-
 \chapter{Little Helpers}
 It is never wrong to have some useful little shortcuts at hand.
 \section{Function Macros}

libtommath_VS2008.vcproj

@@ -392,10 +392,6 @@
 			RelativePath="mp_div_2d.c"
 			>
 		</File>
-		<File
-			RelativePath="mp_div_3.c"
-			>
-		</File>
 		<File
 			RelativePath="mp_div_d.c"
 			>
@@ -760,10 +756,6 @@
 			RelativePath="mp_signed_rsh.c"
 			>
 		</File>
-		<File
-			RelativePath="mp_sqr.c"
-			>
-		</File>
 		<File
 			RelativePath="mp_sqrmod.c"
 			>
@@ -824,6 +816,10 @@
 			RelativePath="s_mp_copy_digs.c"
 			>
 		</File>
+		<File
+			RelativePath="s_mp_div_3.c"
+			>
+		</File>
 		<File
 			RelativePath="s_mp_div_recursive.c"
 			>

makefile

@@ -28,7 +28,7 @@ LCOV_ARGS=--directory .
 #START_INS
 OBJECTS=mp_2expt.o mp_abs.o mp_add.o mp_add_d.o mp_addmod.o mp_and.o mp_clamp.o mp_clear.o mp_clear_multi.o \
 mp_cmp.o mp_cmp_d.o mp_cmp_mag.o mp_cnt_lsb.o mp_complement.o mp_copy.o mp_count_bits.o mp_cutoffs.o \
-mp_div.o mp_div_2.o mp_div_2d.o mp_div_3.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
+mp_div.o mp_div_2.o mp_div_2d.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
 mp_error_to_string.o mp_exch.o mp_expt_u32.o mp_exptmod.o mp_exteuclid.o mp_fread.o mp_from_sbin.o \
 mp_from_ubin.o mp_fwrite.o mp_gcd.o mp_get_double.o mp_get_i32.o mp_get_i64.o mp_get_l.o mp_get_ll.o \
 mp_get_mag_u32.o mp_get_mag_u64.o mp_get_mag_ul.o mp_get_mag_ull.o mp_grow.o mp_init.o mp_init_copy.o \
@@ -42,11 +42,11 @@ mp_prime_strong_lucas_selfridge.o mp_radix_size.o mp_rand.o mp_read_radix.o mp_r
 mp_reduce_2k_l.o mp_reduce_2k_setup.o mp_reduce_2k_setup_l.o mp_reduce_is_2k.o mp_reduce_is_2k_l.o \
 mp_reduce_setup.o mp_root_u32.o mp_rshd.o mp_sbin_size.o mp_set.o mp_set_double.o mp_set_i32.o mp_set_i64.o \
 mp_set_l.o mp_set_ll.o mp_set_u32.o mp_set_u64.o mp_set_ul.o mp_set_ull.o mp_shrink.o mp_signed_rsh.o \
-mp_sqr.o mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
-mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_recursive.o \
-s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o s_mp_invmod.o \
-s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o s_mp_mul.o \
-s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
+mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
+mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_3.o \
+s_mp_div_recursive.o s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o \
+s_mp_invmod.o s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o \
+s_mp_mul.o s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
 s_mp_mul_toom.o s_mp_prime_is_divisible.o s_mp_prime_tab.o s_mp_radix_map.o s_mp_rand_jenkins.o \
 s_mp_rand_platform.o s_mp_sqr.o s_mp_sqr_comba.o s_mp_sqr_karatsuba.o s_mp_sqr_toom.o s_mp_sub.o \
 s_mp_zero_buf.o s_mp_zero_digs.o

makefile.mingw

@@ -30,7 +30,7 @@ LIBMAIN_D =libtommath.dll
 #List of objects to compile (all goes to libtommath.a)
 OBJECTS=mp_2expt.o mp_abs.o mp_add.o mp_add_d.o mp_addmod.o mp_and.o mp_clamp.o mp_clear.o mp_clear_multi.o \
 mp_cmp.o mp_cmp_d.o mp_cmp_mag.o mp_cnt_lsb.o mp_complement.o mp_copy.o mp_count_bits.o mp_cutoffs.o \
-mp_div.o mp_div_2.o mp_div_2d.o mp_div_3.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
+mp_div.o mp_div_2.o mp_div_2d.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
 mp_error_to_string.o mp_exch.o mp_expt_u32.o mp_exptmod.o mp_exteuclid.o mp_fread.o mp_from_sbin.o \
 mp_from_ubin.o mp_fwrite.o mp_gcd.o mp_get_double.o mp_get_i32.o mp_get_i64.o mp_get_l.o mp_get_ll.o \
 mp_get_mag_u32.o mp_get_mag_u64.o mp_get_mag_ul.o mp_get_mag_ull.o mp_grow.o mp_init.o mp_init_copy.o \
@@ -44,11 +44,11 @@ mp_prime_strong_lucas_selfridge.o mp_radix_size.o mp_rand.o mp_read_radix.o mp_r
 mp_reduce_2k_l.o mp_reduce_2k_setup.o mp_reduce_2k_setup_l.o mp_reduce_is_2k.o mp_reduce_is_2k_l.o \
 mp_reduce_setup.o mp_root_u32.o mp_rshd.o mp_sbin_size.o mp_set.o mp_set_double.o mp_set_i32.o mp_set_i64.o \
 mp_set_l.o mp_set_ll.o mp_set_u32.o mp_set_u64.o mp_set_ul.o mp_set_ull.o mp_shrink.o mp_signed_rsh.o \
-mp_sqr.o mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
-mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_recursive.o \
-s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o s_mp_invmod.o \
-s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o s_mp_mul.o \
-s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
+mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
+mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_3.o \
+s_mp_div_recursive.o s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o \
+s_mp_invmod.o s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o \
+s_mp_mul.o s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
 s_mp_mul_toom.o s_mp_prime_is_divisible.o s_mp_prime_tab.o s_mp_radix_map.o s_mp_rand_jenkins.o \
 s_mp_rand_platform.o s_mp_sqr.o s_mp_sqr_comba.o s_mp_sqr_karatsuba.o s_mp_sqr_toom.o s_mp_sub.o \
 s_mp_zero_buf.o s_mp_zero_digs.o

makefile.msvc

@@ -23,7 +23,7 @@ LIBMAIN_S =tommath.lib
 #List of objects to compile (all goes to tommath.lib)
 OBJECTS=mp_2expt.obj mp_abs.obj mp_add.obj mp_add_d.obj mp_addmod.obj mp_and.obj mp_clamp.obj mp_clear.obj mp_clear_multi.obj \
 mp_cmp.obj mp_cmp_d.obj mp_cmp_mag.obj mp_cnt_lsb.obj mp_complement.obj mp_copy.obj mp_count_bits.obj mp_cutoffs.obj \
-mp_div.obj mp_div_2.obj mp_div_2d.obj mp_div_3.obj mp_div_d.obj mp_dr_is_modulus.obj mp_dr_reduce.obj mp_dr_setup.obj \
+mp_div.obj mp_div_2.obj mp_div_2d.obj mp_div_d.obj mp_dr_is_modulus.obj mp_dr_reduce.obj mp_dr_setup.obj \
 mp_error_to_string.obj mp_exch.obj mp_expt_u32.obj mp_exptmod.obj mp_exteuclid.obj mp_fread.obj mp_from_sbin.obj \
 mp_from_ubin.obj mp_fwrite.obj mp_gcd.obj mp_get_double.obj mp_get_i32.obj mp_get_i64.obj mp_get_l.obj mp_get_ll.obj \
 mp_get_mag_u32.obj mp_get_mag_u64.obj mp_get_mag_ul.obj mp_get_mag_ull.obj mp_grow.obj mp_init.obj mp_init_copy.obj \
@@ -37,11 +37,11 @@ mp_prime_strong_lucas_selfridge.obj mp_radix_size.obj mp_rand.obj mp_read_radix.
 mp_reduce_2k_l.obj mp_reduce_2k_setup.obj mp_reduce_2k_setup_l.obj mp_reduce_is_2k.obj mp_reduce_is_2k_l.obj \
 mp_reduce_setup.obj mp_root_u32.obj mp_rshd.obj mp_sbin_size.obj mp_set.obj mp_set_double.obj mp_set_i32.obj mp_set_i64.obj \
 mp_set_l.obj mp_set_ll.obj mp_set_u32.obj mp_set_u64.obj mp_set_ul.obj mp_set_ull.obj mp_shrink.obj mp_signed_rsh.obj \
-mp_sqr.obj mp_sqrmod.obj mp_sqrt.obj mp_sqrtmod_prime.obj mp_sub.obj mp_sub_d.obj mp_submod.obj mp_to_radix.obj mp_to_sbin.obj \
-mp_to_ubin.obj mp_ubin_size.obj mp_unpack.obj mp_xor.obj mp_zero.obj s_mp_add.obj s_mp_copy_digs.obj s_mp_div_recursive.obj \
-s_mp_div_school.obj s_mp_div_small.obj s_mp_exptmod.obj s_mp_exptmod_fast.obj s_mp_get_bit.obj s_mp_invmod.obj \
-s_mp_invmod_odd.obj s_mp_log.obj s_mp_log_d.obj s_mp_log_pow2.obj s_mp_montgomery_reduce_comba.obj s_mp_mul.obj \
-s_mp_mul_balance.obj s_mp_mul_comba.obj s_mp_mul_high.obj s_mp_mul_high_comba.obj s_mp_mul_karatsuba.obj \
+mp_sqrmod.obj mp_sqrt.obj mp_sqrtmod_prime.obj mp_sub.obj mp_sub_d.obj mp_submod.obj mp_to_radix.obj mp_to_sbin.obj \
+mp_to_ubin.obj mp_ubin_size.obj mp_unpack.obj mp_xor.obj mp_zero.obj s_mp_add.obj s_mp_copy_digs.obj s_mp_div_3.obj \
+s_mp_div_recursive.obj s_mp_div_school.obj s_mp_div_small.obj s_mp_exptmod.obj s_mp_exptmod_fast.obj s_mp_get_bit.obj \
+s_mp_invmod.obj s_mp_invmod_odd.obj s_mp_log.obj s_mp_log_d.obj s_mp_log_pow2.obj s_mp_montgomery_reduce_comba.obj \
+s_mp_mul.obj s_mp_mul_balance.obj s_mp_mul_comba.obj s_mp_mul_high.obj s_mp_mul_high_comba.obj s_mp_mul_karatsuba.obj \
 s_mp_mul_toom.obj s_mp_prime_is_divisible.obj s_mp_prime_tab.obj s_mp_radix_map.obj s_mp_rand_jenkins.obj \
 s_mp_rand_platform.obj s_mp_sqr.obj s_mp_sqr_comba.obj s_mp_sqr_karatsuba.obj s_mp_sqr_toom.obj s_mp_sub.obj \
 s_mp_zero_buf.obj s_mp_zero_digs.obj

makefile.shared

@@ -25,7 +25,7 @@ LCOV_ARGS=--directory .libs --directory .
 #START_INS
 OBJECTS=mp_2expt.o mp_abs.o mp_add.o mp_add_d.o mp_addmod.o mp_and.o mp_clamp.o mp_clear.o mp_clear_multi.o \
 mp_cmp.o mp_cmp_d.o mp_cmp_mag.o mp_cnt_lsb.o mp_complement.o mp_copy.o mp_count_bits.o mp_cutoffs.o \
-mp_div.o mp_div_2.o mp_div_2d.o mp_div_3.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
+mp_div.o mp_div_2.o mp_div_2d.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
 mp_error_to_string.o mp_exch.o mp_expt_u32.o mp_exptmod.o mp_exteuclid.o mp_fread.o mp_from_sbin.o \
 mp_from_ubin.o mp_fwrite.o mp_gcd.o mp_get_double.o mp_get_i32.o mp_get_i64.o mp_get_l.o mp_get_ll.o \
 mp_get_mag_u32.o mp_get_mag_u64.o mp_get_mag_ul.o mp_get_mag_ull.o mp_grow.o mp_init.o mp_init_copy.o \
@@ -39,11 +39,11 @@ mp_prime_strong_lucas_selfridge.o mp_radix_size.o mp_rand.o mp_read_radix.o mp_r
 mp_reduce_2k_l.o mp_reduce_2k_setup.o mp_reduce_2k_setup_l.o mp_reduce_is_2k.o mp_reduce_is_2k_l.o \
 mp_reduce_setup.o mp_root_u32.o mp_rshd.o mp_sbin_size.o mp_set.o mp_set_double.o mp_set_i32.o mp_set_i64.o \
 mp_set_l.o mp_set_ll.o mp_set_u32.o mp_set_u64.o mp_set_ul.o mp_set_ull.o mp_shrink.o mp_signed_rsh.o \
-mp_sqr.o mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
-mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_recursive.o \
-s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o s_mp_invmod.o \
-s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o s_mp_mul.o \
-s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
+mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
+mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_3.o \
+s_mp_div_recursive.o s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o \
+s_mp_invmod.o s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o \
+s_mp_mul.o s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
 s_mp_mul_toom.o s_mp_prime_is_divisible.o s_mp_prime_tab.o s_mp_radix_map.o s_mp_rand_jenkins.o \
 s_mp_rand_platform.o s_mp_sqr.o s_mp_sqr_comba.o s_mp_sqr_karatsuba.o s_mp_sqr_toom.o s_mp_sub.o \
 s_mp_zero_buf.o s_mp_zero_digs.o

makefile.unix

@@ -31,7 +31,7 @@ LIBMAIN_S = libtommath.a
 
 OBJECTS=mp_2expt.o mp_abs.o mp_add.o mp_add_d.o mp_addmod.o mp_and.o mp_clamp.o mp_clear.o mp_clear_multi.o \
 mp_cmp.o mp_cmp_d.o mp_cmp_mag.o mp_cnt_lsb.o mp_complement.o mp_copy.o mp_count_bits.o mp_cutoffs.o \
-mp_div.o mp_div_2.o mp_div_2d.o mp_div_3.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
+mp_div.o mp_div_2.o mp_div_2d.o mp_div_d.o mp_dr_is_modulus.o mp_dr_reduce.o mp_dr_setup.o \
 mp_error_to_string.o mp_exch.o mp_expt_u32.o mp_exptmod.o mp_exteuclid.o mp_fread.o mp_from_sbin.o \
 mp_from_ubin.o mp_fwrite.o mp_gcd.o mp_get_double.o mp_get_i32.o mp_get_i64.o mp_get_l.o mp_get_ll.o \
 mp_get_mag_u32.o mp_get_mag_u64.o mp_get_mag_ul.o mp_get_mag_ull.o mp_grow.o mp_init.o mp_init_copy.o \
@@ -45,11 +45,11 @@ mp_prime_strong_lucas_selfridge.o mp_radix_size.o mp_rand.o mp_read_radix.o mp_r
 mp_reduce_2k_l.o mp_reduce_2k_setup.o mp_reduce_2k_setup_l.o mp_reduce_is_2k.o mp_reduce_is_2k_l.o \
 mp_reduce_setup.o mp_root_u32.o mp_rshd.o mp_sbin_size.o mp_set.o mp_set_double.o mp_set_i32.o mp_set_i64.o \
 mp_set_l.o mp_set_ll.o mp_set_u32.o mp_set_u64.o mp_set_ul.o mp_set_ull.o mp_shrink.o mp_signed_rsh.o \
-mp_sqr.o mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
-mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_recursive.o \
-s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o s_mp_invmod.o \
-s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o s_mp_mul.o \
-s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
+mp_sqrmod.o mp_sqrt.o mp_sqrtmod_prime.o mp_sub.o mp_sub_d.o mp_submod.o mp_to_radix.o mp_to_sbin.o \
+mp_to_ubin.o mp_ubin_size.o mp_unpack.o mp_xor.o mp_zero.o s_mp_add.o s_mp_copy_digs.o s_mp_div_3.o \
+s_mp_div_recursive.o s_mp_div_school.o s_mp_div_small.o s_mp_exptmod.o s_mp_exptmod_fast.o s_mp_get_bit.o \
+s_mp_invmod.o s_mp_invmod_odd.o s_mp_log.o s_mp_log_d.o s_mp_log_pow2.o s_mp_montgomery_reduce_comba.o \
+s_mp_mul.o s_mp_mul_balance.o s_mp_mul_comba.o s_mp_mul_high.o s_mp_mul_high_comba.o s_mp_mul_karatsuba.o \
 s_mp_mul_toom.o s_mp_prime_is_divisible.o s_mp_prime_tab.o s_mp_radix_map.o s_mp_rand_jenkins.o \
 s_mp_rand_platform.o s_mp_sqr.o s_mp_sqr_comba.o s_mp_sqr_karatsuba.o s_mp_sqr_toom.o s_mp_sub.o \
 s_mp_zero_buf.o s_mp_zero_digs.o

mp_div_d.c

@@ -28,23 +28,26 @@ mp_err mp_div_d(const mp_int *a, mp_digit b, mp_int *c, mp_digit *d)
    }
 
    /* power of two ? */
-   if ((b & (b - 1u)) == 0u) {
+   if (MP_HAS(MP_DIV_2) && (b == 2u)) {
+      if (d != NULL) {
+         *d = mp_isodd(a) ? 1u : 0u;
+      }
+      return (c == NULL) ? MP_OKAY : mp_div_2(a, c);
+   }
+   if (MP_HAS(MP_DIV_2D) && MP_IS_2EXPT(b)) {
       ix = 1;
       while ((ix < MP_DIGIT_BIT) && (b != (((mp_digit)1)<<ix))) {
          ix++;
       }
       if (d != NULL) {
          *d = a->dp[0] & (((mp_digit)1<<(mp_digit)ix) - 1uL);
       }
-      if (c != NULL) {
-         return mp_div_2d(a, ix, c, NULL);
-      }
-      return MP_OKAY;
+      return (c == NULL) ? MP_OKAY : mp_div_2d(a, ix, c, NULL);
    }
 
    /* three? */
-   if (MP_HAS(MP_DIV_3) && (b == 3u)) {
-      return mp_div_3(a, c, d);
+   if (MP_HAS(S_MP_DIV_3) && (b == 3u)) {
+      return s_mp_div_3(a, c, d);
    }
 
    /* no easy answer [c'est la vie].  Just division */

mp_log_u32.c

@@ -17,7 +17,7 @@ mp_err mp_log_u32(const mp_int *a, uint32_t base, uint32_t *c)
       return MP_VAL;
    }
 
-   if (MP_HAS(S_MP_LOG_POW2) && ((base & (base - 1u)) == 0u)) {
+   if (MP_HAS(S_MP_LOG_POW2) && MP_IS_2EXPT(base)) {
       *c = s_mp_log_pow2(a, base);
       return MP_OKAY;
    }