Closed Bug 593850 Opened 15 years ago Closed 15 years ago

fix a bunch of failing webgl tests

Categories

(Core :: Graphics: CanvasWebGL, defect)

defect
Not set
normal

Tracking

()

RESOLVED FIXED

People

(Reporter: vlad, Assigned: vlad)

Details

Attachments

(6 files)

There's a bunch of webgl tests that are failing for simple reasons.
this is already checked in upstream; these tests shouldn't be referencing debugcontext.
Attachment #472438 - Flags: review?(bjacob)
Attachment #472434 - Flags: review?(bjacob) → review+
Comment on attachment 472435 [details] [diff] [review] fix gl-scissor test Note: need to send these changes upstream or they will be lost the next time we sync.
Attachment #472435 - Flags: review?(bjacob) → review+
Attachment #472436 - Flags: review?(bjacob) → review+
Attachment #472437 - Flags: review?(bjacob) → review+
Attachment #472438 - Flags: review?(bjacob) → review+
Comment on attachment 472439 [details] [diff] [review] fix bad-args and null-args tests NS_IMETHODIMP WebGLContext::LinkProgram(nsIWebGLProgram *pobj) { + if (!pobj) + return ErrorInvalidValue("linkProgram"); + GLuint progname; WebGLProgram *program; if (!GetConcreteObjectAndGLName("linkProgram", pobj, &program, &progname)) I don't understand this one. GetConcreteObject and friends already do generate ErrorInvalidValue on null objects (unless &isNull is passed). So why is it needed to add these if(!pobj) checks?
Also, just "linkProgram" is too elusive an error message. How about "linkProgram: null program passed".
(In reply to comment #7) > Comment on attachment 472435 [details] [diff] [review] > fix gl-scissor test > > Note: need to send these changes upstream or they will be lost the next time we > sync. Already checked in upstream (In reply to comment #8) > Comment on attachment 472439 [details] [diff] [review] > fix bad-args and null-args tests > > NS_IMETHODIMP > WebGLContext::LinkProgram(nsIWebGLProgram *pobj) > { > + if (!pobj) > + return ErrorInvalidValue("linkProgram"); > + > GLuint progname; > WebGLProgram *program; > if (!GetConcreteObjectAndGLName("linkProgram", pobj, &program, &progname)) > > > I don't understand this one. GetConcreteObject and friends already do generate > ErrorInvalidValue on null objects (unless &isNull is passed). So why is it > needed to add these if(!pobj) checks? Ah yeah, you're right.. this was added in an earlier version where I thought these were supposed to be no-ops, but I was wrong. I'll get rid of them.
all checked in, with ErrorInvalidValue crap removed.
Status: NEW → RESOLVED
Closed: 15 years ago
Resolution: --- → FIXED
Comment on attachment 472439 [details] [diff] [review] fix bad-args and null-args tests This patch has already landed.
Attachment #472439 - Flags: review?(bjacob)
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: