Update NSS to NSS_3_15_BETA2.

Update NSS to NSS_3_15_BETA2.

R=agl@chromium.org,rsleevi@chromium.org
BUG=233732
TEST=no build errors on iOS, Mac OS X, and Windows.

[wtc, 2013-04-23 00:58:05]

https://codereview.chromium.org/13898013/diff/1/nss/lib/freebl/ecl/ecp_256_32.c
File nss/lib/freebl/ecl/ecp_256_32.c (left):

https://codereview.chromium.org/13898013/diff/1/nss/lib/freebl/ecl/ecp_256_32...
nss/lib/freebl/ecl/ecp_256_32.c:1368: PORT_Assert(MP_USED(n) * sizeof(mp_digit)
<= 32);

In the NSS upstream, it is difficult for this file to use
PORT_Assert, which is why this PORT_Assert was removed.

Do you want to keep this PORT_Assert in our copy of this file?

https://codereview.chromium.org/13898013/diff/1/nss/lib/ssl/sslerr.h
File nss/lib/ssl/sslerr.h (right):

https://codereview.chromium.org/13898013/diff/1/nss/lib/ssl/sslerr.h#newcode191
nss/lib/ssl/sslerr.h:191: SSL_ERROR_RX_UNEXPECTED_CERT_STATUS     =
(SSL_ERROR_BASE + 125),

There is another copy of this file (sslerr.h) in
src/net/third_party/nss/ssl. I verified that the value of
SSL_ERROR_RX_UNEXPECTED_CERT_STATUS is the same in both
copies of this file.

https://codereview.chromium.org/13898013/diff/1/nss/lib/util/hasht.h
File nss/lib/util/hasht.h (right):

https://codereview.chromium.org/13898013/diff/1/nss/lib/util/hasht.h#newcode9
nss/lib/util/hasht.h:9: #include "prtypes.h"

I will remove this from the NSS upstream. I don't see any
NSPR types used in this header file.

https://codereview.chromium.org/13898013/diff/1/nss/lib/util/seccomon.h
File nss/lib/util/seccomon.h (right):

https://codereview.chromium.org/13898013/diff/1/nss/lib/util/seccomon.h#newco...
nss/lib/util/seccomon.h:63: unsigned int len;

I think this field should be renamed |nitems| or |count|.

|len| is too similar to the |len| field of SECItem (see line
56 above).

[agl, 2013-04-24 22:27:00]

LGTM.

https://codereview.chromium.org/13898013/diff/1/nss/lib/freebl/ecl/ecp_256_32.c
File nss/lib/freebl/ecl/ecp_256_32.c (left):

https://codereview.chromium.org/13898013/diff/1/nss/lib/freebl/ecl/ecp_256_32...
nss/lib/freebl/ecl/ecp_256_32.c:1368: PORT_Assert(MP_USED(n) * sizeof(mp_digit)
<= 32);
On 2013/04/23 00:58:05, wtc wrote:
> 
> In the NSS upstream, it is difficult for this file to use
> PORT_Assert, which is why this PORT_Assert was removed.
> 
> Do you want to keep this PORT_Assert in our copy of this file?

I think this is fine.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/freebl/ecl/ecl.c
File nss/lib/freebl/ecl/ecl.c (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/freebl/ecl/ecl.c#n...
nss/lib/freebl/ecl/ecl.c:267: break;
I'm slightly surprised not to see something about P-384 here.

[wtc, 2013-04-24 22:49:45]

These are notes to myself. You may ignore them.

I will make these changes in the NSS upstream.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certdb/genname.c
File nss/lib/certdb/genname.c (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certdb/genname.c#n...
nss/lib/certdb/genname.c:1468: current =
CERT_GetNextNameConstraint((CERTNameConstraint*)current);

I verified this "const cast" is fine.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certhigh/ocsp.c
File nss/lib/certhigh/ocsp.c (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certhigh/ocsp.c#ne...
nss/lib/certhigh/ocsp.c:144: cert_DupOCSPCertID(CERTOCSPCertID *src);
The argument should be 'const'.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certhigh/ocsp.c#ne...
nss/lib/certhigh/ocsp.c:787: *certIDWasConsumed = PR_FALSE;
We should fix the indentation.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certhigh/ocsp.c#ne...
nss/lib/certhigh/ocsp.c:827: OCSP_TRACE(("Not caching response because the
response is not newer than the cache"));
Fold this long line.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certhigh/ocsp.c#ne...
nss/lib/certhigh/ocsp.c:1798: if (!dest)
Fix indentation.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/certhigh/ocsp.c#ne...
nss/lib/certhigh/ocsp.c:1805: goto loser;     \
Add curly braces. Combine nested ifs using &&.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/pkcs7/p7decode.c
File nss/lib/pkcs7/p7decode.c (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/pkcs7/p7decode.c#n...
nss/lib/pkcs7/p7decode.c:1763: NULL, HASH_AlgNULL, keepcerts, 0);
This 0 should be NULL.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/pkcs7/p7decode.c#n...
nss/lib/pkcs7/p7decode.c:1872: NULL, HASH_AlgNULL, PR_FALSE, 0);
This 0 should be NULL.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c
File nss/lib/util/secitem.c (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c#new...
nss/lib/util/secitem.c:356: * If array is not NULL, the above has set
array->data and
This comment needs to be updated to say array->items, etc.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c#new...
nss/lib/util/secitem.c:363: void secitem_FreeArray(SECItemArray *array, PRBool
zero_items, PRBool freeit)
Mark this static.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c#new...
nss/lib/util/secitem.c:370: for (i=0; i<array->len; ++i) {
Add spaces around operators.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c#new...
nss/lib/util/secitem.c:381: 
BUG: array->items also needs to be freed.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c#new...
nss/lib/util/secitem.c:402: if (!from || !from->items || !from->len)
I think it is legal for from->items to be NULL.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.c#new...
nss/lib/util/secitem.c:409: for (i=0; i<from->len; ++i) {
Add spaces around operators.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.h
File nss/lib/util/secitem.h (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/util/secitem.h#new...
nss/lib/util/secitem.h:96: extern SECItemArray *SECITEM_DupArray(PLArenaPool
*arena, const SECItemArray *from);
Fold the long line.

[wtc, 2013-04-24 22:51:13]

agl: thanks for the review.

https://codereview.chromium.org/13898013/diff/3008/nss/lib/freebl/ecl/ecl.c
File nss/lib/freebl/ecl/ecl.c (right):

https://codereview.chromium.org/13898013/diff/3008/nss/lib/freebl/ecl/ecl.c#n...
nss/lib/freebl/ecl/ecl.c:267: break;

On 2013/04/24 22:27:00, agl wrote:
> I'm slightly surprised not to see something about P-384 here.

Indeed. I also noticed this problem (that the P-384 code
is not being used) and filed a bug report:
https://bugzilla.mozilla.org/show_bug.cgi?id=841901

[commit-bot: I haz the power, 2013-04-24 23:28:40]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/wtc@chromium.org/13898013/3008

[commit-bot: I haz the power, 2013-04-24 23:29:38]

Change committed as 196262

[wtc, 2013-05-04 01:17:57]

https://codereview.chromium.org/13898013/diff/1/nss/lib/util/hasht.h
File nss/lib/util/hasht.h (right):

https://codereview.chromium.org/13898013/diff/1/nss/lib/util/hasht.h#newcode9
nss/lib/util/hasht.h:9: #include "prtypes.h"

On 2013/04/23 00:58:05, wtc wrote:
> 
> I will remove this from the NSS upstream. I don't see any
> NSPR types used in this header file.

I was wrong. This header uses the PRBool type.