Fix ASTC Decode_astc_8x8_unorm_array_7 testcase segfault

[walcht]

As partially discussed here: #1173.

I skimmed through some tests and by grepping for "KTX_SUCCESS" all checks almost always use EXPECT_XX which should, probably, be changed to ASSERT_XX (because why proceed to a segfault/other failure when a core Ktx function returned an exit error code?).

[MarkCallow]

After I merge this, you will no longer be a "first-time contributor" so pushes thereafter to PR #1167 will start a CI run. Is #1167 ready for that?

[MarkCallow]

I skimmed through some tests and by grepping for "KTX_SUCCESS" all checks almost always use EXPECT_XX which should, probably, be changed to ASSERT_XX (because why proceed to a segfault/other failure when a core Ktx function returned an exit error code?).

Only the KTX_SUCCESS checks on functions that will return a nullptr on failure, e.g. ktxTexture?_Create*, need to be changed. For those cases where the following nullptr check outputs a message, the message must be moved to the KTX_SUCCESS check.

Please make these changes, if you have time.

[walcht]

So I looked at other KTX_SUCCESS checks, and they all seem to make sense. Some are:

EXPECT_EQ(result, KTX_SUCCESS);
ASSERT_TRUE(texture != NULL) << "ktxTexture?_Create failed: " << ktxErrorString(result);

While others:

ASSERT_TRUE(result == KTX_SUCCESS);
ASSERT_TRUE(texture != NULL) << "ktxTexture_CreateFromMemory failed: " << ktxErrorString(result);

Some others do not assert on texture != nullptr but have reasonable if statements afterwards:

if (texture) { /* other stuff/tests */ }

So I just changed the single check that is causing the issue without touching anything else.

Is #1167 ready for that?

Finally, yes.

[MarkCallow]

ASSERT_TRUE(result == KTX_SUCCESS);
ASSERT_TRUE(texture != NULL) << "ktxTexture_CreateFromMemory failed: " << ktxErrorString(result);

IIUC, in this case the test will stop if result != KTX_SUCCESS which means the error string for result will not be printed. Please change these to EXPECT_EQ.

[MarkCallow]

Please merge the latest main which fixes the cloudflare-related Windows CI failure.

[MarkCallow]

Please merge latest main here and in PR #1167.

[MarkCallow]

I'll come back in an hour and if you've done it, I'll run the builds.