793373 - B2G SMS: xpcshell test script test_ril_worker_sms.js fails

Status: RESOLVED WONTFIX

(Core :: DOM: Device Interfaces, defect)

Description

[Vicamo Yang [:vicamo][:vyang]]

test_ril_worker_sms.js has been broken for a long time. The first breakage, which can be fixed with a simple patch, was introduced in bug 712933 due to `message.body` being renamed to `message.fullBody`. The second major breakage was introduced in bug 744709 by bringing in Settings API into RadioInterfaceLayer's constructor.

Comment 1

[Vicamo Yang [:vicamo][:vyang]]

Attachment: error 1

Comment 2

[Vicamo Yang [:vicamo][:vyang]]

Attachment: error 2

Comment 3

[Vicamo Yang [:vicamo][:vyang]]

Attachment: fix for error 1

Comment 4

[Vicamo Yang [:vicamo][:vyang]]

Attachment: Part 1: s/body/fullBody/

Comment 5

[Vicamo Yang [:vicamo][:vyang]]

Attachment: Part 2: Move SMS related functions out of RadioInterfaceLayer

Comment 6

[Vicamo Yang [:vicamo][:vyang]]

Attachment: Part 3: fix test_ril_worker_sms.js test failed due to getting Settings service in RadioInterfaceLayer

try: https://tbpl.mozilla.org/?tree=Try&rev=22dc6aaf20d6

Comment 7

[Vicamo Yang [:vicamo][:vyang]]

We'll have to fix this to enable SMS parser test cases.

Comment 8

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Comment on attachment 681367 [details] [diff] [review]
Part 3: fix test_ril_worker_sms.js test failed due to getting Settings service in RadioInterfaceLayer

Review of attachment 681367 [details] [diff] [review]:
-----------------------------------------------------------------

Yes, test passed! Thank you.

Comment 9

[Lawrence Mandel [:lmandel] (use needinfo)]

basecamp-. Please request Aurora approval to uplift.

Comment 10

[Mozilla RelEng Bot]

Try run for 22dc6aaf20d6 is complete.
Detailed breakdown of the results available here:
    https://tbpl.mozilla.org/?tree=Try&rev=22dc6aaf20d6
Results (out of 306 total builds):
    success: 292
    warnings: 12
    failure: 2
Builds (or logs if builds failed) available at:
http://ftp.mozilla.org/pub/mozilla.org/firefox/try-builds/[e-mail]-22dc6aaf20d6

Comment 11

[Vicamo Yang [:vicamo][:vyang]]

Attachment: Part 2: don't call ctor of RIL in xpcshell tests

Sorry, I forget its existence at all :S

A super-tiny fix comes to my mind after bug 833697. We can simply skip the constructor of RadioInterfaceLayer and preserve everything other than the test case itself.

Comment 12

[Hsin-Yi Tsai (she/her) [:hsinyi]]

Comment on attachment 717024 [details] [diff] [review]
Part 2: don't call ctor of RIL in xpcshell tests

Review of attachment 717024 [details] [diff] [review]:
-----------------------------------------------------------------

Great catch!

Comment 13

[Mozilla RelEng Bot]

Try run for 728a7669d105 is complete.
Detailed breakdown of the results available here:
    https://tbpl.mozilla.org/?tree=Try&rev=728a7669d105
Results (out of 18 total builds):
    exception: 18
Builds (or logs if builds failed) available at:
http://ftp.mozilla.org/pub/mozilla.org/firefox/try-builds/[e-mail]-728a7669d105

Comment 14

[Vicamo Yang [:vicamo][:vyang]]

Reset platform to x86_64 linux because this bug only happens when we run B2G xpcshell test cases in desktop build.

Comment 15

[Vicamo Yang [:vicamo][:vyang]]

(In reply to Mozilla RelEng Bot from comment #13)
> Try run for 728a7669d105 is complete.
> Results (out of 18 total builds):
>     exception: 18

Cancelled manually. A new run is available in https://tbpl.mozilla.org/?tree=Try&rev=55adef2232f7

Comment 16

[Vicamo Yang [:vicamo][:vyang]]

Apparently somehow this breaks B2G ARM xpcshell tests. Since we have bug 775281, easy of use |test.sh xpcshell| script, been landed, this bug is no longer meaningful.  Mark as WONTFIX.