Mark dashing_gpu and complexclip2_rect_aa_gpu as ignore-failure on Win7 perspective

expectations/gm/Test-Win7-ShuttleA-HD2000-x86-Release/expected-results.json

 {
   "expected-results": {
     "3x3bitmaprect_565.png": {
       "allowed-digests": [
         [
           "bitmap-64bitMD5", 
           9785004568147979273
         ]
       ], 
       "bugs": [
@@ skipping 3772 matching lines @@
       ], 
       "reviewed-by-human": false
     }, 
     "complexclip2_rect_aa_gpu.png": {
       "allowed-digests": [
         [
           "bitmap-64bitMD5", 
           3908297226363946093
         ]
       ], 
+      "ignore-failure": true,

[epoger, 2014/03/24 18:36:02]

It worries me to mark results as ignore-failure without a tracking bug (as in
this case) or associated with a bug which is already marked as fixed (as for
dashing_gpu.png).

Maybe associate these with bug 2245, and keep it open (low priority) to track?

       "reviewed-by-human": true
     }, 
     "complexclip2_rect_aa_pdf-poppler.png": {
       "allowed-digests": [
         [
           "bitmap-64bitMD5", 
           17155654729794872162
         ]
       ], 
       "bugs": [
@@ skipping 702 matching lines @@
     }, 
     "dashing_gpu.png": {
       "allowed-digests": [
         [
           "bitmap-64bitMD5", 
           13624796742030499023
         ]
       ], 
       "bugs": [
         2146
-      ], 
+      ],
+      "ignore-failure": true,
       "reviewed-by-human": false
     }, 
     "dashing_pdf-poppler.png": {
       "allowed-digests": [
         [
           "bitmap-64bitMD5", 
           13521151190286408436
         ]
       ], 
       "bugs": [
@@ skipping 8551 matching lines @@
           "bitmap-64bitMD5", 
           15843716056234692659
         ]
       ], 
       "bugs": [
         2146
       ], 
       "reviewed-by-human": false
     }
   }
-}
+}