934831 - [gonk-misc]Checking jb4.2.2 support

Status: RESOLVED FIXED

(Firefox OS Graveyard :: General, defect)

Description

[thomas tsai]

Attachment: 0001-No-fakeappops-in-jb4.2.2.patch

fakeappops is not supported in jb4.2.2,but jb4.3. There are still some OEM/ODM developing in jb4.2.2, so this check can pass the build.

Comment 1

[thomas tsai]

Attachment: jb4.2.2.patch

fakesched and fakeappops are not supported in jb4.2.2

Comment 2

[Kai-Zhen Li [:seinlin][:kli]]

Attachment: jb4.2.2 support

Comment 3

[Marco Chen [:mchen]]

Comment on attachment 827239 [details] [review]
jb4.2.2 support

Since the SDK version can be used to distinguish ICS/JB MR1/MR2 already, we can remove additional "folder checking".

Comment 4

[Kai-Zhen Li [:seinlin][:kli]]

Comment on attachment 827239 [details] [review]
jb4.2.2 support

Marco, the folder checking is remove and pull request is updated. Thanks!

Comment 5

[Marco Chen [:mchen]]

Comment on attachment 827239 [details] [review]
jb4.2.2 support

ICS build will be broken so please exclude the ICS too. Thanks.

Comment 6

[Michael Wu [:mwu]]

If possible, please check the source tree to determine whether things things need to be built, rather than checking the platform version.

fakesched was originally written for 4.2. Are you sure it isn't required?

Comment 7

[Kai-Zhen Li [:seinlin][:kli]]

Comment on attachment 827239 [details] [review]
jb4.2.2 support

The pull request is updated,

-fakesched
4.2 and 4.3 get different function defined for "requestPriority".
determine by using PLATFORM_SDK_VERSION

-fakeappops
both 4.2 and 4.3 get this dir "frameworks/native/libs/binder"
determine by using "frameworks/native/libs/binder/IAppOpsService.cpp"

Comment 8

[Marco Chen [:mchen]]

Comment on attachment 827239 [details] [review]
jb4.2.2 support

It looks good to me already so transfer to reviewer - :mwu.

Hi Michael,

Please help to review this patch . Thanks.

Comment 9

[Michael Wu [:mwu]]

Thanks, this looks better.

Just one thing I'd like to check - does it work if you implement both versions requestPriority instead of having a compile time check? The linker should just pick up the right one.

Comment 10

[Kai-Zhen Li [:seinlin][:kli]]

Overloading the function "requestPriority" also does work. So is it better to have both versions of "requestPriority" instead of having a compile  flags?

Comment 11

[Michael Wu [:mwu]]

I think so - it's usually worth avoiding ifdefs.

Comment 12

[Kai-Zhen Li [:seinlin][:kli]]

Pull request is updated.

Comment 13

[Michael Wu [:mwu]]

Merged