Re: [PATCH v5 2/8] lib/mpi: Extend the MPI library

[Tianjia Zhang]

On 2020/7/10 21:12, Marcelo Henrique Cerri wrote:
> Hi, Tianjia.
>
> On Thu, Jul 09, 2020 at 04:40:09PM +0800, Tianjia Zhang wrote:
> > Expand the mpi library based on libgcrypt, and the ECC algorithm of
> > mpi based on libgcrypt requires these functions.
> > Some other algorithms will be developed based on mpi ecc, such as SM2.
> >
> > Signed-off-by: Tianjia Zhang <tianjia.zhang@xxxxxxxxxxxxxxxxx>
> > ---
> >   include/linux/mpi.h    |  88 +++++++++++
> >   lib/mpi/Makefile       |   5 +
> >   lib/mpi/mpi-add.c      | 207 +++++++++++++++++++++++++
> >   lib/mpi/mpi-bit.c      | 251 ++++++++++++++++++++++++++++++
> >   lib/mpi/mpi-cmp.c      |  46 ++++--
> >   lib/mpi/mpi-div.c      | 238 +++++++++++++++++++++++++++++
> >   lib/mpi/mpi-internal.h |  53 +++++++
> >   lib/mpi/mpi-inv.c      | 143 ++++++++++++++++++
> >   lib/mpi/mpi-mod.c      | 155 +++++++++++++++++++
> >   lib/mpi/mpi-mul.c      |  94 ++++++++++++
> >   lib/mpi/mpicoder.c     | 336 +++++++++++++++++++++++++++++++++++++++++
> >   lib/mpi/mpih-div.c     | 294 ++++++++++++++++++++++++++++++++++++
> >   lib/mpi/mpih-mul.c     |  25 +++
> >   lib/mpi/mpiutil.c      | 204 +++++++++++++++++++++++++
> >   14 files changed, 2129 insertions(+), 10 deletions(-)
> >   create mode 100644 lib/mpi/mpi-add.c
> >   create mode 100644 lib/mpi/mpi-div.c
> >   create mode 100644 lib/mpi/mpi-inv.c
> >   create mode 100644 lib/mpi/mpi-mod.c
> >   create mode 100644 lib/mpi/mpi-mul.c
> >
> > diff --git a/lib/mpi/mpi-add.c b/lib/mpi/mpi-add.c
> > new file mode 100644
> > index 000000000000..9afad7832737
> > --- /dev/null
> > +++ b/lib/mpi/mpi-add.c
> > @@ -0,0 +1,207 @@
> > +/* mpi-add.c  -  MPI functions
> > + * Copyright (C) 1994, 1996, 1998, 2001, 2002,
> > + *               2003 Free Software Foundation, Inc.
> > + *
> > + * This file is part of Libgcrypt.
> > + *
> > + * Note: This code is heavily based on the GNU MP Library.
> > + *	 Actually it's the same code with only minor changes in the
> > + *	 way the data is stored; this is to support the abstraction
> > + *	 of an optional secure memory allocation which may be used
> > + *	 to avoid revealing of sensitive data due to paging etc.
> > + */
> > +
> > +#include "mpi-internal.h"
> > +
> > +/****************
> > + * Add the unsigned integer V to the mpi-integer U and store the
> > + * result in W. U and V may be the same.
> > + */
> > +void mpi_add_ui(MPI w, MPI u, unsigned long v)
> > +{
> > +	mpi_ptr_t wp, up;
> > +	mpi_size_t usize, wsize;
> > +	int usign, wsign;
> > +
> > +	usize = u->nlimbs;
> > +	usign = u->sign;
> > +	wsign = 0;
> > +
> > +	/* If not space for W (and possible carry), increase space.  */
> > +	wsize = usize + 1;
> > +	if (w->alloced < wsize)
> > +		mpi_resize(w, wsize);
>
> You are ignoring the mpi_resize() return. I believe these new functions
> need to return an int to indicate errors as mpi_powm() does.

Yes, of course.  Thanks for pointing it out.

Thanks,
Tianjia

> > +
> > +	/* These must be after realloc (U may be the same as W).  */
> > +	up = u->d;
> > +	wp = w->d;
> > +
> > +	if (!usize) {  /* simple */
> > +		wp[0] = v;
> > +		wsize = v ? 1:0;
> > +	} else if (!usign) {  /* mpi is not negative */
> > +		mpi_limb_t cy;
> > +		cy = mpihelp_add_1(wp, up, usize, v);
> > +		wp[usize] = cy;
> > +		wsize = usize + cy;
> > +	} else {
> > +		/* The signs are different.  Need exact comparison to determine
> > +		 * which operand to subtract from which.
> > +		 */
> > +		if (usize == 1 && up[0] < v) {
> > +			wp[0] = v - up[0];
> > +			wsize = 1;
> > +		} else {
> > +			mpihelp_sub_1(wp, up, usize, v);
> > +			/* Size can decrease with at most one limb. */
> > +			wsize = usize - (wp[usize-1] == 0);
> > +			wsign = 1;
> > +		}
> > +	}
> > +
> > +	w->nlimbs = wsize;
> > +	w->sign   = wsign;
> > +}
> > +