New issue
Advanced search Search tips
Note: Color blocks (like or ) mean that a user may not be available. Tooltip shows the reason.
Starred by 24 users

Issue metadata

Status: WontFix
Owner:
OOO until 3-4
Closed: Jun 2015
Components:
EstimatedDays: ----
NextAction: ----
OS: Windows
Pri: 3
Type: Bug



Sign in to add a comment
link

Issue 167951: Build NSS in 64-bit on Windows

Reported by rsleevi@chromium.org, Jan 2 2013 Project Member

Issue description

NSS's freebl has optimized assembly when built on Windows 64-bit, through the use of MASM.

nss.gyp should support building these when building for Win x64.
 

Comment 1 by rsleevi@chromium.org, Jan 2 2013

Blocking: chromium:8606

Comment 2 by wtc@chromium.org, Jan 6 2013

Cc: wtc@chromium.org
Owner: rsleevi@chromium.org
Status: Started
rsleevi has written a CL: https://codereview.chromium.org/11738002/

Comment 3 by bugdroid1@chromium.org, Feb 14 2013

Project Member
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=182578

------------------------------------------------------------------------
r182578 | wtc@chromium.org | 2013-02-14T23:45:32.926186Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/deps/third_party/nss/nss.gyp?r1=182578&r2=182577&pathrev=182578

Define _AMD64_, MP_CHAR_STORE_SLOW, and MP_IS_LITTLE_ENDIAN
for 64-bit Windows builds.

This matches the upstream NSS build system.

R=jschuh@chromium.org,rsleevi@chromium.org
BUG= 167951 
TEST=none

Review URL: https://chromiumcodereview.appspot.com/12210134
------------------------------------------------------------------------

Comment 4 by bugdroid1@chromium.org, Feb 15 2013

Project Member
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=182723

------------------------------------------------------------------------
r182723 | wtc@chromium.org | 2013-02-15T16:08:50.368295Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/DEPS?r1=182723&r2=182722&pathrev=182723

Update nss_revision to r182578 to pick up WIN64 build
system tweaks and a fix for invalid read in rc4_wordconv.

The underlying CLs are:
https://codereview.chromium.org/12210134/
https://codereview.chromium.org/12226071/

R=rsleevi@chromium.org
BUG= 167951 , 174140 
TEST=none 

Review URL: https://chromiumcodereview.appspot.com/12259038
------------------------------------------------------------------------

Comment 5 by bugdroid1@chromium.org, Feb 20 2013

Project Member
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=183330

------------------------------------------------------------------------
r183330 | simonjam@chromium.org | 2013-02-19T23:55:39.181378Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/DEPS?r1=183330&r2=183329&pathrev=183330

Revert 182723 - Speculative. Will reland if no perf change.

> Update nss_revision to r182578 to pick up WIN64 build
> system tweaks and a fix for invalid read in rc4_wordconv.
> 
> The underlying CLs are:
> https://codereview.chromium.org/12210134/
> https://codereview.chromium.org/12226071/
> 
> R=rsleevi@chromium.org
> BUG= 167951 , 174140 
> TEST=none 
> 
> Review URL: https://chromiumcodereview.appspot.com/12259038

TBR=wtc@chromium.org

Review URL: https://chromiumcodereview.appspot.com/12303041
------------------------------------------------------------------------

Comment 6 by bugdroid1@chromium.org, Feb 20 2013

Project Member
The following revision refers to this bug:
    http://src.chromium.org/viewvc/chrome?view=rev&revision=183354

------------------------------------------------------------------------
r183354 | simonjam@chromium.org | 2013-02-20T01:17:42.313860Z

Changed paths:
   M http://src.chromium.org/viewvc/chrome/trunk/src/DEPS?r1=183354&r2=183353&pathrev=183354

Revert 183330 - Speculation was wrong. Relanding original.
> Revert 182723 - Speculative. Will reland if no perf change.
> 
> > Update nss_revision to r182578 to pick up WIN64 build
> > system tweaks and a fix for invalid read in rc4_wordconv.
> > 
> > The underlying CLs are:
> > https://codereview.chromium.org/12210134/
> > https://codereview.chromium.org/12226071/
> > 
> > R=rsleevi@chromium.org
> > BUG= 167951 , 174140 
> > TEST=none 
> > 
> > Review URL: https://chromiumcodereview.appspot.com/12259038
> 
> TBR=wtc@chromium.org
> 
> Review URL: https://chromiumcodereview.appspot.com/12303041

TBR=wtc@chromium.org
Review URL: https://codereview.chromium.org/12311002
------------------------------------------------------------------------

Comment 7 by jsc...@chromium.org, Mar 3 2013

Looks like this might be fixed now?

Comment 8 by rsleevi@chromium.org, Mar 4 2013

The optimizations (eg: ASM) are not in place yet.

Comment 9 by bugdroid1@chromium.org, Mar 10 2013

Project Member
Labels: -Area-Internals -Internals-Network-SSL Cr-Internals Cr-Internals-Network-SSL

Comment 10 by stan...@gmail.com, Mar 21 2014

Any progress here?

Comment 11 by jsc...@chromium.org, Apr 14 2014

rsleevi@ - Anything else left to do here or are we good? Also, should this block the Win64 launch (at the moment I would think not)?

Comment 12 by rsleevi@chromium.org, Apr 14 2014

jschuh: NSS builds on Win64, but we still don't have any of the optimized assembly. We were shuffling about upstream NSS making the assembly depend on YASM instead of MASM, but that has largely stalled over objections from Red Hat about adopting YASM wholesale (currently, both Chrome and Mozilla Firefox treat YASM as a build dependency, and thus have it available).

Porting the MASM code to YASM is still an exercise that upstream would support for Windows, but it will mean we'll continue to maintain parallel assembly (for !Win and Win).

Comment 13 by jsc...@chromium.org, Apr 14 2014

Blocking: -chromium:8606

Comment 14 by laforge@google.com, Apr 28 2015

Cc: -wtc@chromium.org

Comment 15 by rsleevi@chromium.org, Jun 10 2015

Status: WontFix
Marking this as WontFix.

As per comment #12, we have it building, but w/o optimizations. However, Windows has transitioned to BoringSSL, where we do have optimizations. Problem solved.

Sign in to add a comment