gm refactor: extract run_multiple_configs() and run_multiple_modes()

gm refactor: extract run_multiple_configs() and run_multiple_modes()
Committed: https://code.google.com/p/skia/source/detail?r=8269

[epoger, 2013-03-20 17:26:42]

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp
File gm/gmmain.cpp (right):

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1
gm/gmmain.cpp:1: /*
Refactor only... no functional change.

Originally I tried to do this as patchset 2 of
https://codereview.chromium.org/12825005 ('gm self-test: make sure we report
failures in individual rendering modes'), but there have been many changes to
gmmain over the past 2 days and merging them together proved quite difficult.

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1150
gm/gmmain.cpp:1150: ErrorBitfield run_multiple_configs(GMMain &gmmain, GM *gm,
const SkTDArray<size_t> &configs,
All of this code just extracted from tools_main().

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1272
gm/gmmain.cpp:1272: ErrorBitfield run_multiple_modes(GMMain &gmmain, GM *gm,
const ConfigData &compareConfig,
All of this code just extracted from tools_main().

[epoger, 2013-03-20 17:28:29]

note to self: in "sk red" on c128.i

[scroggo, 2013-03-20 17:41:30]

lgtm. What is the motivation for these changes? it looks like run_multiple_* are
each only called once atm.

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp
File gm/gmmain.cpp (right):

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1
gm/gmmain.cpp:1: /*
On 2013/03/20 17:26:42, epoger wrote:
> Refactor only... no functional change.
> 
> Originally I tried to do this as patchset 2 of
> https://codereview.chromium.org/12825005 ('gm self-test: make sure we report
> failures in individual rendering modes'), but there have been many changes to
> gmmain over the past 2 days and merging them together proved quite difficult.

Sorry about that!

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1318
gm/gmmain.cpp:1318: gm, compareConfig, "-serialize", bitmap, &comparisonBitmap);
Over 100 lines.

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1339
gm/gmmain.cpp:1339: gm, kRTree_BbhType,
SkPicture::kUsePathBoundsForClip_RecordingFlag);
100 lines.

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1357
gm/gmmain.cpp:1357: gm, kTileGrid_BbhType,
SkPicture::kUsePathBoundsForClip_RecordingFlag,
100 lines.

https://codereview.chromium.org/12536007/diff/1/gm/tests/run.sh
File gm/tests/run.sh (right):

https://codereview.chromium.org/12536007/diff/1/gm/tests/run.sh#newcode8
gm/tests/run.sh:8: # To generate new baselines when gm behavior changes, run
gm/tests/rebaseline.sh
+1

[epoger, 2013-03-20 17:54:38]

On 2013/03/20 17:41:30, scroggo wrote:
> lgtm. What is the motivation for these changes? it looks like run_multiple_*
are
> each only called once atm.

Discussed live... this refactor will make it easier to reorder those operations
(e.g., test modes BEFORE testing configs) or make related changes.

[epoger, 2013-03-20 17:54:49]

PTAL at patchset 2.

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp
File gm/gmmain.cpp (right):

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1
gm/gmmain.cpp:1: /*
On 2013/03/20 17:41:30, scroggo wrote:
> On 2013/03/20 17:26:42, epoger wrote:
> > Refactor only... no functional change.
> > 
> > Originally I tried to do this as patchset 2 of
> > https://codereview.chromium.org/12825005 ('gm self-test: make sure we report
> > failures in individual rendering modes'), but there have been many changes
to
> > gmmain over the past 2 days and merging them together proved quite
difficult.
> 
> Sorry about that!

Hey, it's progress... a good thing.

https://codereview.chromium.org/12536007/diff/1/gm/gmmain.cpp#newcode1318
gm/gmmain.cpp:1318: gm, compareConfig, "-serialize", bitmap, &comparisonBitmap);
On 2013/03/20 17:41:30, scroggo wrote:
> Over 100 lines.

Fixed this and other line length / style stuff in patchset 2.

[scroggo, 2013-03-20 17:59:01]

lgtm

[epoger, 2013-03-20 17:59:32]

Committed patchset #2 manually as r8269 (presubmit successful).