Chromium Code Reviews
chromiumcodereview-hr@appspot.gserviceaccount.com (chromiumcodereview-hr) | Please choose your nickname with Settings | Help | Chromium Project | Gerrit Changes | Sign out
(418)

Issue 10273021: Revive Tag() function (Closed)

Created:
8 years, 7 months ago by bashi
Modified:
8 years, 7 months ago
Reviewers:
Yusuke Sato, agl
CC:
chromium-reviews
Visibility:
Public.

Description

Removed TAG() macro and revived Tag() function for big endian platform. BUG=124629 TEST=ran test_{un,}malicious_fonts.sh

Patch Set 1 #

Total comments: 2

Patch Set 2 : #

Unified diffs Side-by-side diffs Delta from patch set Stats (+42 lines, -43 lines) Patch
M src/ots.cc View 1 6 chunks +42 lines, -43 lines 0 comments Download

Messages

Total messages: 6 (0 generated)
bashi
Hi yusuke-san, Could you take a look?
8 years, 7 months ago (2012-05-01 00:38:51 UTC) #1
Yusuke Sato
https://chromiumcodereview.appspot.com/10273021/diff/1/src/ots.cc File src/ots.cc (right): https://chromiumcodereview.appspot.com/10273021/diff/1/src/ots.cc#newcode86 src/ots.cc:86: // Try endianness checks with zero, one and two ...
8 years, 7 months ago (2012-05-01 02:23:31 UTC) #2
bashi
http://codereview.chromium.org/10273021/diff/1/src/ots.cc File src/ots.cc (right): http://codereview.chromium.org/10273021/diff/1/src/ots.cc#newcode86 src/ots.cc:86: // Try endianness checks with zero, one and two ...
8 years, 7 months ago (2012-05-01 04:05:48 UTC) #3
Yusuke Sato
lgtm, but please wait for comments from agl and jfkthame. On 2012/05/01 04:05:48, bashik wrote: ...
8 years, 7 months ago (2012-05-02 06:22:20 UTC) #4
jfkthame
Sorry, I haven't had an opportunity to test this on various big-endian systems, but both ...
8 years, 7 months ago (2012-05-14 15:26:31 UTC) #5
bashi
8 years, 7 months ago (2012-05-14 23:52:28 UTC) #6
Ok, I'm going to submit the CL. Please let me know if there is any problems.
Thank you for the help.

On 2012/05/14 15:26:31, jfkthame wrote:
> Sorry, I haven't had an opportunity to test this on various big-endian
systems,
> but both Cameron Kaiser (who maintains TenFourFox for PPC) and I expect that
it
> will be fine. I'd suggest you go ahead with this change in OTS, as the current
> #ifdef solution is definitely broken on some platforms.
> 
> On 2012/05/02 06:22:20, Yusuke Sato wrote:
> > lgtm, but please wait for comments from agl and jfkthame.
> > 
> > On 2012/05/01 04:05:48, bashik wrote:
> > > http://codereview.chromium.org/10273021/diff/1/src/ots.cc
> > > File src/ots.cc (right):
> > > 
> > > http://codereview.chromium.org/10273021/diff/1/src/ots.cc#newcode86
> > > src/ots.cc:86: // Try endianness checks with zero, one and two leading
> > > underscores,
> > > On 2012/05/01 02:23:31, Yusuke Sato wrote:
> > > > The #ifdefs look too complex to me. What about reviving Tag() and then
> using
> > > the
> > > > following simple solution?
> > > > 
> > > > const struct {
> > > >   const char* tag;
> > > > ...
> > > > } table_parsers[] = {
> > > >   { "maxp", ...
> > > > 
> > > > ...
> > > > 
> > > > if (Tag(table_parsers[i].tag) == Tag("head")) {
> > > > 
> > > 
> > > Done. Thank you for suggestion!

Powered by Google App Engine
This is Rietveld 408576698