| Message ID | 20150824145214.26239.98271.stgit@tstruk-mobl1 (mailing list archive) |
|---|---|
| State | Accepted |
| Delegated to: | Herbert Xu |
| Headers | show |
On Mon, Aug 24, 2015 at 07:52:14AM -0700, Tadeusz Struk wrote: > Change mpi_read_buffer to return a number without leading zeros > so that mpi_read_buffer and mpi_get_buffer return the same thing. > > Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com> Applied.
Tadeusz Struk <tadeusz.struk@intel.com> wrote: > Change mpi_read_buffer to return a number without leading zeros > so that mpi_read_buffer and mpi_get_buffer return the same thing. Hmmm... This would appear to have the undesirable side effect of causing 1 out of 256 module signing keys to be unusable (ie. those that begin with 00). David -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
David Howells <dhowells@redhat.com> wrote: > > Change mpi_read_buffer to return a number without leading zeros > > so that mpi_read_buffer and mpi_get_buffer return the same thing. > > Hmmm... This would appear to have the undesirable side effect of causing 1 > out of 256 module signing keys to be unusable (ie. those that begin with 00). At least, I think it's this, but it's not entirely clear since it's not been reproduced yet. David -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
On Tue, Sep 15, 2015 at 10:09:56AM +0100, David Howells wrote: > Tadeusz Struk <tadeusz.struk@intel.com> wrote: > > > Change mpi_read_buffer to return a number without leading zeros > > so that mpi_read_buffer and mpi_get_buffer return the same thing. > > Hmmm... This would appear to have the undesirable side effect of causing 1 > out of 256 module signing keys to be unusable (ie. those that begin with 00). I thought Tadeusz was just restoring the original behaviour, no? Thanks,
On 09/15/2015 05:24 AM, Herbert Xu wrote: >>> Change mpi_read_buffer to return a number without leading zeros >>> > > so that mpi_read_buffer and mpi_get_buffer return the same thing. >> > >> > Hmmm... This would appear to have the undesirable side effect of causing 1 >> > out of 256 module signing keys to be unusable (ie. those that begin with 00). > I thought Tadeusz was just restoring the original behaviour, no? That was the intention. David, I can ran some more test if you have the key that was failing and be willing to share it. -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html
diff --git a/lib/mpi/mpicoder.c b/lib/mpi/mpicoder.c index bc0a1da..95c52a9 100644 --- a/lib/mpi/mpicoder.c +++ b/lib/mpi/mpicoder.c @@ -146,18 +146,25 @@ int mpi_read_buffer(MPI a, uint8_t *buf, unsigned buf_len, unsigned *nbytes, uint8_t *p; mpi_limb_t alimb; unsigned int n = mpi_get_size(a); - int i; + int i, lzeros = 0; - if (buf_len < n || !buf) + if (buf_len < n || !buf || !nbytes) return -EINVAL; if (sign) *sign = a->sign; - if (nbytes) - *nbytes = n; + p = (void *)&a->d[a->nlimbs] - 1; + + for (i = a->nlimbs * sizeof(alimb) - 1; i >= 0; i--, p--) { + if (!*p) + lzeros++; + else + break; + } p = buf; + *nbytes = n - lzeros; for (i = a->nlimbs - 1; i >= 0; i--) { alimb = a->d[i]; @@ -178,6 +185,19 @@ int mpi_read_buffer(MPI a, uint8_t *buf, unsigned buf_len, unsigned *nbytes, #else #error please implement for this limb size. #endif + + if (lzeros > 0) { + if (lzeros >= sizeof(alimb)) { + p -= sizeof(alimb); + } else { + mpi_limb_t *limb1 = (void *)p - sizeof(alimb); + mpi_limb_t *limb2 = (void *)p - sizeof(alimb) + + lzeros; + *limb1 = *limb2; + p -= lzeros; + } + lzeros -= sizeof(alimb); + } } return 0; } @@ -197,7 +217,7 @@ EXPORT_SYMBOL_GPL(mpi_read_buffer); */ void *mpi_get_buffer(MPI a, unsigned *nbytes, int *sign) { - uint8_t *buf, *p; + uint8_t *buf; unsigned int n; int ret; @@ -220,14 +240,6 @@ void *mpi_get_buffer(MPI a, unsigned *nbytes, int *sign) kfree(buf); return NULL; } - - /* this is sub-optimal but we need to do the shift operation - * because the caller has to free the returned buffer */ - for (p = buf; !*p && *nbytes; p++, --*nbytes) - ; - if (p != buf) - memmove(buf, p, *nbytes); - return buf; } EXPORT_SYMBOL_GPL(mpi_get_buffer);
Change mpi_read_buffer to return a number without leading zeros so that mpi_read_buffer and mpi_get_buffer return the same thing. Signed-off-by: Tadeusz Struk <tadeusz.struk@intel.com> --- lib/mpi/mpicoder.c | 38 +++++++++++++++++++++++++------------- 1 file changed, 25 insertions(+), 13 deletions(-) -- To unsubscribe from this list: send the line "unsubscribe linux-crypto" in the body of a message to majordomo@vger.kernel.org More majordomo info at http://vger.kernel.org/majordomo-info.html