Commit a78324d9 authored by Andy Polyakov's avatar Andy Polyakov
Browse files

bn/bn_add.c: address performance regression. 



Performance regression was reported for EC key generation between
1.0.2 and 1.1.x [in GH#2891]. It naturally depends on platform,
values between 6 and 9% were observed.

Reviewed-by: default avatarRichard Levitte <levitte@openssl.org>
(Merged from https://github.com/openssl/openssl/pull/4743)
parent 8e4ec5b2

crypto/bn/asm/x86_64-gcc.c

+8 −6
Original line number Diff line number Diff line
@@ -225,9 +225,10 @@ BN_ULONG bn_add_words(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, 
                  "       adcq    (%5,%2,8),%0    \n"

                  "       movq    %0,(%3,%2,8)    \n"

                  "       lea     1(%2),%2        \n"

                  "       loop    1b              \n"

                  "       sbbq    %0,%0           \n":"=&r" (ret), "+c"(n),

                  "+r"(i)

                  "       dec     %1              \n"

                  "       jnz     1b              \n"

                  "       sbbq    %0,%0           \n"

                  :"=&r" (ret), "+c"(n), "+r"(i)

                  :"r"(rp), "r"(ap), "r"(bp)

                  :"cc", "memory");
@@ -251,9 +252,10 @@ BN_ULONG bn_sub_words(BN_ULONG *rp, const BN_ULONG *ap, const BN_ULONG *bp, 
                  "       sbbq    (%5,%2,8),%0    \n"

                  "       movq    %0,(%3,%2,8)    \n"

                  "       lea     1(%2),%2        \n"

                  "       loop    1b              \n"

                  "       sbbq    %0,%0           \n":"=&r" (ret), "+c"(n),

                  "+r"(i)

                  "       dec     %1              \n"

                  "       jnz     1b              \n"

                  "       sbbq    %0,%0           \n"

                  :"=&r" (ret), "+c"(n), "+r"(i)

                  :"r"(rp), "r"(ap), "r"(bp)

                  :"cc", "memory");

crypto/bn/bn_add.c

+5 −1
Original line number Diff line number Diff line
@@ -141,9 +141,13 @@ int BN_usub(BIGNUM *r, const BIGNUM *a, const BIGNUM *b) 
        borrow &= (t1 == 0);

    }



    while (max && *--rp == 0)

        max--;



    r->top = max;

    r->neg = 0;

    bn_correct_top(r);

    bn_pollute(r);



    return 1;

}