OpenSSL: fix serial number tests.

OpenSSL: fix serial number tests.

OpenSSL internally performs the removal of the leading zero bytes that we
previously did ourselves. Since the X509Certificate API now specifies that the
serial number should be returned in ASN.1 form, we have to put the leading zero
byte back for positive numbers with the MSB set.

BUG=none
TEST=net_unittests

[joth, 2011-10-31 17:17:45]

LGTM thanks!
2 very optional comments...

http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_opens...
File net/base/x509_certificate_openssl.cc (right):

http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_opens...
net/base/x509_certificate_openssl.cc:337: num->data[0] >= 0x80) {
nit: I find the last term slightly easier to read as
&& num->data[0] & 0x80

but I don't care that much either way.

http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_opens...
net/base/x509_certificate_openssl.cc:339: // byte in order that the MSB isn't
set.
is there a reasonable hint you could add here as to why this is only needed on
openssl? maybe useful if anyone is ever diffing this with nss etc.

[agl, 2011-10-31 17:23:45]

Cheers. Will land once the tryservers look happy.

http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_opens...
File net/base/x509_certificate_openssl.cc (right):

http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_opens...
net/base/x509_certificate_openssl.cc:337: num->data[0] >= 0x80) {
On 2011/10/31 17:17:45, joth wrote:
> nit: I find the last term slightly easier to read as
> && num->data[0] & 0x80
> 
> but I don't care that much either way.

Done.

http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_opens...
net/base/x509_certificate_openssl.cc:339: // byte in order that the MSB isn't
set.
On 2011/10/31 17:17:45, joth wrote:
> is there a reasonable hint you could add here as to why this is only needed on
> openssl? maybe useful if anyone is ever diffing this with nss etc.

Have expanded the comment.

[joth, 2011-10-31 17:54:21]

thanks,

On 31 October 2011 17:23, <agl@chromium.org> wrote:

> Cheers. Will land once the tryservers look happy.
>
>
>
> http://codereview.chromium.**org/8387032/diff/1/net/base/**
>
x509_certificate_openssl.cc<http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_openssl.cc>
> File net/base/x509_certificate_**openssl.cc (right):
>
> http://codereview.chromium.**org/8387032/diff/1/net/base/**
>
x509_certificate_openssl.cc#**newcode337<http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_openssl.cc#newcode337>
> net/base/x509_certificate_**openssl.cc:337: num->data[0] >= 0x80) {
> On 2011/10/31 17:17:45, joth wrote:
>
>> nit: I find the last term slightly easier to read as
>> && num->data[0] & 0x80
>>
>
>  but I don't care that much either way.
>>
>
> Done.
>
>
> http://codereview.chromium.**org/8387032/diff/1/net/base/**
>
x509_certificate_openssl.cc#**newcode339<http://codereview.chromium.org/8387032/diff/1/net/base/x509_certificate_openssl.cc#newcode339>
> net/base/x509_certificate_**openssl.cc:339: // byte in order that the MSB
> isn't set.
> On 2011/10/31 17:17:45, joth wrote:
>
>> is there a reasonable hint you could add here as to why this is only
>>
> needed on
>
>> openssl? maybe useful if anyone is ever diffing this with nss etc.
>>
>
> Have expanded the comment.
>
>
http://codereview.chromium.**org/8387032/<http://codereview.chromium.org/8387...
>