Re: Accelerating exact rooting work

2013-05-04 Thread Boris Zbarsky
On 5/4/13 3:43 PM, Ehsan Akhgari wrote: As far as I can tell they seem to map to mozilla::ErrorResult values being used in some contexts. Should we just ignore them? Yes. ErrorResult has a union which sometimes contains a JS::Value (which it then manually roots) and the analysis was confused

Re: Accelerating exact rooting work

2013-05-04 Thread Tom Schuster
You can ignore unnecessary roots. They don't hurt us right now and we can fix them after everything else is done. On Sat, May 4, 2013 at 9:43 PM, Ehsan Akhgari wrote: > There are a ton of messages in that list of the following form: > > Function 'foo' has unnecessary root 'bar' in 'baz.cpp:123' >

Re: Accelerating exact rooting work

2013-05-04 Thread Ehsan Akhgari
There are a ton of messages in that list of the following form: Function 'foo' has unnecessary root 'bar' in 'baz.cpp:123' As far as I can tell they seem to map to mozilla::ErrorResult values being used in some contexts. Should we just ignore them? Ehsan On 2013-05-04 12:56 PM, Boris Zbarsk

Re: Accelerating exact rooting work

2013-05-04 Thread Boris Zbarsky
On 4/23/13 11:18 AM, Ehsan Akhgari wrote: Can the stuff in objdir/dom/bindings be fixed whole-sale by changing the WebIDL codegen? This is nearly done: we're down to the issues in bug 868715, which largely affect the js-implemented codegen at the moment, and the problem of what to do with Typ

Re: Accelerating exact rooting work

2013-04-23 Thread Andrew McCreight
- Original Message - > Does this also apply to code holding on to JSObject*s? I've been > adding a whole bunch of those cases lately. If you are telling the cycle collector about those JSObjects, then it should get fixed whenever CCed stuff is fixed. Andrew _

Re: Accelerating exact rooting work

2013-04-23 Thread Boris Zbarsky
On 4/23/13 11:18 AM, Ehsan Akhgari wrote: Does this also apply to code holding on to JSObject*s? If you're holding them on the heap... sigh. Just trace them and use fromMarkedLocation as needed; there is nothing better at the moment. You can't use JS::Rooted on the heap. Can the stuff in

Re: Accelerating exact rooting work

2013-04-23 Thread smaug
On 04/23/2013 06:23 PM, Ehsan Akhgari wrote: Hmm, another question. Your list includes a bunch of stuff under tools/profiler, and I took a quick look and picked JSObjectBuilder.cpp. Changing the JS::Value's there to JS::RootedValue's cause compiler errors about conversion from jsvals when us

Re: Accelerating exact rooting work

2013-04-23 Thread Kyle Huey
On Tue, Apr 23, 2013 at 8:18 AM, Ehsan Akhgari wrote: > Can the stuff in objdir/dom/bindings be fixed whole-sale by changing the > WebIDL codegen? > Yes. bz is looking into it. - Kyle ___ dev-platform mailing list dev-platform@lists.mozilla.org https:

Re: Accelerating exact rooting work

2013-04-23 Thread Ehsan Akhgari
Hmm, another question. Your list includes a bunch of stuff under tools/profiler, and I took a quick look and picked JSObjectBuilder.cpp. Changing the JS::Value's there to JS::RootedValue's cause compiler errors about conversion from jsvals when using macros such as INT_TO_JSVAL, STRING_TO_JSV

Re: Accelerating exact rooting work

2013-04-23 Thread Ehsan Akhgari
On 2013-04-23 9:07 AM, Tom Schuster wrote: At the moment it's really just Jono working full time on this, and terrence and other people reviewing. This stuff is actually quite easy and you can expect really fast review times from our side. In some parts of the code rooting could literally just m

Re: Accelerating exact rooting work

2013-04-23 Thread Kevin Gadd
Related nitpick: > *Warning:* The information at SpiderMonkey Garbage Collection > Tipsand > in the JSAPI User Guideis woefully out of date and

Re: Accelerating exact rooting work

2013-04-23 Thread Till Schneidereit
There's a rooting guide on the wiki here: https://developer.mozilla.org/en-US/docs/SpiderMonkey/GC_Rooting_Guide It's not very thorough, but it's something. On Tue, Apr 23, 2013 at 3:14 PM, Tom Schuster wrote: > You wont have to bother with HandleObject or RootedObject etc. These > are interna

Re: Accelerating exact rooting work

2013-04-23 Thread Tom Schuster
You wont have to bother with HandleObject or RootedObject etc. These are internal to js/ and we take care of these cases. Just use JS::Handle or JS::Rooted. Thanks for pointing people to this file. PS: The correct handle for Jon Coppeard is of course jonco! On Tue, Apr 23, 2013 at 3:58 PM, smaug

Re: Accelerating exact rooting work

2013-04-23 Thread smaug
On 04/23/2013 04:07 PM, Tom Schuster wrote: At the moment it's really just Jono working full time on this, and terrence and other people reviewing. This stuff is actually quite easy and you can expect really fast review times from our side. In some parts of the code rooting could literally just

Re: Accelerating exact rooting work

2013-04-23 Thread Tom Schuster
At the moment it's really just Jono working full time on this, and terrence and other people reviewing. This stuff is actually quite easy and you can expect really fast review times from our side. In some parts of the code rooting could literally just mean to replace JS::Value to JS::RootedValue a

Accelerating exact rooting work

2013-04-22 Thread Robert O'Callahan
On Tue, Apr 23, 2013 at 5:36 AM, Terrence Cole wrote: > Our exact rooting work is at a spot right now where we could easily use > more hands to accelerate the process. The main problem is that the work > is easy and tedious: a hard sell for pretty much any hacker at mozilla. > It sounds worthwhi