Summary: | Change default setting of run-javascriptcore-tests not to build JSC by default | ||||||
---|---|---|---|---|---|---|---|
Product: | WebKit | Reporter: | Peter Molnar <pmolnar.u-szeged> | ||||
Component: | Tools / Tests | Assignee: | Nobody <webkit-unassigned> | ||||
Status: | RESOLVED WONTFIX | ||||||
Severity: | Normal | CC: | ap, dbates, dfarler, eric, fpizlo, ggaren, lforschler, rniwa | ||||
Priority: | P2 | ||||||
Version: | 528+ (Nightly build) | ||||||
Hardware: | Unspecified | ||||||
OS: | Unspecified | ||||||
Attachments: |
|
Description
Peter Molnar
2014-01-16 02:58:45 PST
Created attachment 221353 [details]
patch
The Apple webkit testers pass the --no-build flag to the script already. So, I think this will not cause any problems. Phil, will this affect your JSC bots? (In reply to comment #0) > This script is ran most of the time on buildbots, in these cases there's no need to build JSC, as it is built right before running this script. > When running these tests manually on a device (like dev boards where we only run tests, and we build with cross-compile), it's quite an annoyance to specify the --no-build option every time we do that. > > It may be wise to change this behavior, to make the script build JSC only if the --build option is specified explicitly. This will cover those few cases when we want to run the build from this script. For me, I always use run-javascriptcore-tests to build JSC and I never use build-jsc directly. This change would be a huge annoyance for me. I can also see this leading to problems if someone reruns tests after making changes but forgets to pass --build. I don't like that prospect. Since this appears to be an annoyance to one person no matter what the default is, I think the wisest thing to do is to not change anything. |