Closed Bug 1023877 Opened 10 years ago Closed 10 years ago

Reference workload does not work on Flame

Categories

(Firefox OS Graveyard :: Performance, defect, P1)

ARM
Gonk (Firefox OS)
defect

Tracking

(b2g-v2.0 fixed, b2g-v2.1 fixed)

RESOLVED FIXED
2.0 S5 (4july)
Tracking Status
b2g-v2.0 --- fixed
b2g-v2.1 --- fixed

People

(Reporter: davehunt, Assigned: jhylands)

References

Details

(Keywords: perf, Whiteboard: [c=automation p=1 s= u=])

Attachments

(1 file)

The reference workload tries to push to /sdcard which does not exist on Flame. It should be updated to push to /storage/sdcard0 or /storage/sdcard1 when the target is Flame.
Assignee: nobody → jhylands
Status: NEW → ASSIGNED
Component: General → Performance
Keywords: perf
Priority: -- → P1
Whiteboard: [c=automation p=1 s= u=]
Attached file Pull Request for fix
Attachment #8442205 - Flags: review?(dave.hunt)
Comment on attachment 8442205 [details]
Pull Request for fix

I haven't tested, but this looks good to me. We can probably remove the commented out lines too.

If we have another device in the future with a different storage path then we'll need to add that too. A more robust solution would be nice, but we shouldn't block on that.
Attachment #8442205 - Flags: review?(dave.hunt) → review+
I'll test this on Hamachi as requested by Jon on IRC. If it works, I'll merge it.
Flags: needinfo?(dave.hunt)
Comment on attachment 8442205 [details]
Pull Request for fix

After testing, it appears this is not working 100%. The gallery workload assumes the DCIM/100MZLLA path will exist in the storage location, however it may not.

I also think we should exit if any workload fails to push. At the moment we ignore any failures and exit with a code of 0. This means that in continuous integration we may miss important issues. Let me know if you'd like me to raise this as a separate bug.
Attachment #8442205 - Flags: review+ → review-
Flags: needinfo?(dave.hunt)
Lets definitely raise that as a separate bug. I've updated the PR to fix the gallery issue.
Comment on attachment 8442205 [details]
Pull Request for fix

Looks good. I'll merge this once Try/Travis are happy.
Attachment #8442205 - Flags: review- → review+
Flags: needinfo?(dave.hunt)
(In reply to Jon Hylands [:jhylands] from comment #5)
> Lets definitely raise that as a separate bug.

I've raised bug 1028261.
Flags: needinfo?(dave.hunt)
Target Milestone: --- → 2.0 S4 (20june)
Landed in:
https://github.com/mozilla-b2g/gaia/commit/75b7384aca898ce177be17023d2dc4392e9f10dd
Status: ASSIGNED → RESOLVED
Closed: 10 years ago
Resolution: --- → FIXED
Target Milestone: 2.0 S4 (20june) → 2.0 S5 (4july)
Comment on attachment 8442205 [details]
Pull Request for fix

The tests on Datazilla are failing to install the reference-workload-light on the Flame on 2.0 because it is installing to the incorrect SD card location. We need this uplifted to 2.0 so we can run these tests correctly against a Flame.

[Bug caused by] (feature/regressing bug #): 1023877
[User impact] if declined: n/a
[Testing completed]: yes
[Risk to taking this patch] (and alternatives if risky): low since this affects workload scripts for testing, not user facing
[String changes made]: n/a
Attachment #8442205 - Flags: approval-gaia-v2.0?
Attachment #8442205 - Flags: approval-gaia-v2.0? → approval-gaia-v2.0+
You need to log in before you can comment on or make changes to this bug.

Attachment

General

Created:
Updated:
Size: