Remove prtypes.h and prcpucfg*.h from base/third_party/nspr.

Remove prtypes.h and prcpucfg*.h from base/third_party/nspr.

Copy the type and macro definitions we need to base/third_party/nspr/prtime.h.

Add chromium-prtypes.h to crypto/third_party/nss.

content/common/plugin_list_posix.cc doesn't need to include prcpucfg_linux.h.

R=ananta@chromium.org,mark@chromium.org,rsleevi@chromium.org

Committed: https://src.chromium.org/viewvc/chrome?view=rev&revision=259148

[wtc, 2014-03-22 15:50:19]

Ananta: you only need to review content/, but you can also review base/ since
you worked on prtime.cc before.

Mark, Ryan: please review the entire CL. Thanks.

The main goal is to eliminate the prcpucfg*.h files. This also prepares for the
consolidation of crypto code on OpenSSL. At that time, only
base/third_party/nspr/prtime.{h,cc} will remain.

The typedefs we need are conveniently identified by compilation errors. I
checked all the relevant files manually for the macros they need.

https://codereview.chromium.org/209343003/diff/60001/base/third_party/nspr/pr...
File base/third_party/nspr/prtime.cc (left):

https://codereview.chromium.org/209343003/diff/60001/base/third_party/nspr/pr...
base/third_party/nspr/prtime.cc:446: #if defined(XP_MAC)

This is obsolete code. XP_MAC refers to Mac OS Classic.

https://codereview.chromium.org/209343003/diff/60001/base/third_party/nspr/pr...
File base/third_party/nspr/prtypes.h (left):

https://codereview.chromium.org/209343003/diff/60001/base/third_party/nspr/pr...
base/third_party/nspr/prtypes.h:94: #define PR_IMPORT(__type) extern __type

NSPR_API(__type) expands to either PR_EXPORT(__type) or PR_IMPORT(__type), both
of which expand to extern __type. This fact is used in the definition of
NSPR_API in prtime.h.

https://codereview.chromium.org/209343003/diff/60001/content/common/plugin_li...
File content/common/plugin_list_posix.cc (left):

https://codereview.chromium.org/209343003/diff/60001/content/common/plugin_li...
content/common/plugin_list_posix.cc:36: #include
"base/third_party/nspr/prcpucfg_linux.h"

This header is not needed. I don't know why it is here. Perhaps some old code
needed it.

https://codereview.chromium.org/209343003/diff/60001/crypto/crypto.gypi
File crypto/crypto.gypi (right):

https://codereview.chromium.org/209343003/diff/60001/crypto/crypto.gypi#newco...
crypto/crypto.gypi:19: 'third_party/nss/chromium-blapit.h',

I moved chromium-blapi.h and chromium-blapit.h here because hmac_win.cc includes
chromium-blapi.h, which includes chromium-blapit.h.

[ananta, 2014-03-24 18:54:24]

lgtm

[Ryan Sleevi, 2014-03-24 19:14:08]

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
crypto/third_party/nss/chromium-prtypes.h:25: #endif
This one needs a bit more justification/explanation. This isn't guaranteed,
AFAIK, so I suspect you're relying on some implementation-specific/dependent
aspects that need to be documented.

eg: If this is only for NaCl, why aren't we always defining LINUX, since AIUI,
the nacl toolchain emulates a POSIX API?

[Ryan Sleevi, 2014-03-24 19:14:09]

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
crypto/third_party/nss/chromium-prtypes.h:25: #endif
This one needs a bit more justification/explanation. This isn't guaranteed,
AFAIK, so I suspect you're relying on some implementation-specific/dependent
aspects that need to be documented.

eg: If this is only for NaCl, why aren't we always defining LINUX, since AIUI,
the nacl toolchain emulates a POSIX API?

[Mark Mentovai, 2014-03-24 20:11:24]

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
crypto/third_party/nss/chromium-prtypes.h:25: #endif
Ryan Sleevi wrote:
> This one needs a bit more justification/explanation. This isn't guaranteed,
> AFAIK, so I suspect you're relying on some implementation-specific/dependent
> aspects that need to be documented.

gcc and clang define __SIZEOF_LONG__. Maybe that would be helpful.

This macro is only used in sha512.cc. Another alternative would be to just do
something more direct in that file (GG_UINT64_C?). The ULLC macro in that file
is marginally frightening.

[wtc, 2014-03-24 21:55:10]

Mark, Ryan: please review patch set 6.

I decided to use the same MPL/GPL/LGPL triple license in chromium-prtypes.h.
Although I wrote everything in chromium-prtypes.h, it is best to avoid licensing
questions by using the same license.

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/60001/crypto/third_party/nss/c...
crypto/third_party/nss/chromium-prtypes.h:25: #endif

On 2014/03/24 19:14:10, Ryan Sleevi wrote:
> This one needs a bit more justification/explanation. This isn't guaranteed,
> AFAIK, so I suspect you're relying on some implementation-specific/dependent
> aspects that need to be documented.

Done.
 
> eg: If this is only for NaCl, why aren't we always defining LINUX, since AIUI,
> the nacl toolchain emulates a POSIX API?

This header is used when USE_OPENSSL is not defined, so it is not only for NaCl.

On 2014/03/24 20:11:25, Mark Mentovai wrote:
>
> gcc and clang define __SIZEOF_LONG__. Maybe that would be helpful.

Thanks. I also considered testing __LP64__. Since 'long' is always 32 bits on
Windows, testing a gcc and clang specific macro is not a problem here.

https://codereview.chromium.org/209343003/diff/100001/crypto/third_party/nss/...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/100001/crypto/third_party/nss/...
crypto/third_party/nss/chromium-prtypes.h:58: #if LONG_MAX > 2147483647L

I replaced INT_MAX with 2147483647L.

I used INT_MAX originally because I wasn't sure what type the C preprocessor
uses for integer constants in arithmetic expressions. I just looked it up. The
spec says "[Each] integer constant is considered to be suffixed with L, so that
all arithmetic is taken to be long or unsigned long." Since long is at least 32
bits, 2147483647L won't overflow.

[Ryan Sleevi, 2014-03-25 00:07:08]

lgtm

https://codereview.chromium.org/209343003/diff/100001/crypto/third_party/nss/...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/100001/crypto/third_party/nss/...
crypto/third_party/nss/chromium-prtypes.h:2: * Version: MPL 1.1/GPL 2.0/LGPL 2.1
This is derived from MPL 2.0 code, is it not?
http://mxr.mozilla.org/nspr/source/LICENSE

[wtc, 2014-03-25 00:23:04]

https://codereview.chromium.org/209343003/diff/100001/crypto/third_party/nss/...
File crypto/third_party/nss/chromium-prtypes.h (right):

https://codereview.chromium.org/209343003/diff/100001/crypto/third_party/nss/...
crypto/third_party/nss/chromium-prtypes.h:2: * Version: MPL 1.1/GPL 2.0/LGPL 2.1

On 2014/03/25 00:07:08, Ryan Sleevi wrote:
> This is derived from MPL 2.0 code, is it not?
> http://mxr.mozilla.org/nspr/source/LICENSE

I wrote everything in this file (except for the PR_MIN macro) from scratch. This
is why I originally used Chromium's BSD license. But now I think all the files
related to sha512.cc in this directory should have the same license. Those files
were copied when NSS was using the triple license.

prtypes.h didn't change much between the triple license and MPL 2.0, so we can
consider this file as being derived from the triple license version of
prtypes.h.

[wtc, 2014-03-25 00:24:58]

The CQ bit was checked by wtc@chromium.org

[commit-bot: I haz the power, 2014-03-25 00:25:03]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/wtc@chromium.org/209343003/100001

[Mark Mentovai, 2014-03-25 00:55:08]

LGTM

[commit-bot: I haz the power, 2014-03-25 01:00:54]

The CQ bit was unchecked by commit-bot@chromium.org

[commit-bot: I haz the power, 2014-03-25 01:00:55]

Retried try job too often on chromium_presubmit for step(s) presubmit
http://build.chromium.org/p/tryserver.chromium/buildstatus?builder=chromium_p...

[wtc, 2014-03-25 02:09:04]

The CQ bit was checked by wtc@chromium.org

[commit-bot: I haz the power, 2014-03-25 02:09:54]

CQ is trying da patch. Follow status at
https://chromium-status.appspot.com/cq/wtc@chromium.org/209343003/100001

[commit-bot: I haz the power, 2014-03-25 04:28:34]

Change committed as 259148