Hi Colin, The biggest problem is that Chrome doesn't give ANY stacktrace, nor does it allow for breaking on error. I've tried compiling with 'pretty' so I could breakpoint on error and figure out the offending code, but it won't work.
We are using GWT 2.8-beta1, but I have the same issue using 2.7. On Monday, March 28, 2016 at 8:55:35 AM UTC-4, Colin Alworth wrote: > > The only issue I know of that causes this is > https://github.com/gwtproject/gwt/issues/8233 (and also apparently > https://github.com/gwtproject/gwt/issues/8229), which is fixed in GWT > 2.7. This issue was caused by invoking a method with many arguments, rather > than wrapping those arguments in a list and invoking with the list itself. > It is possible that you haven't had enough items in the list to cause this, > but that Chrome has changed the size of its stack (or some other > implementation detail) and you are now tickling it. > > Just a guess, but if it is the case, upgrading will solve it. If not, is > it possible to get a stack trace out of the error? > > On Mon, Mar 28, 2016 at 5:01 AM 'Marius Gerwinn' via GWT Contributors < > [email protected] <javascript:>> wrote: > >> Hi Josh, >> >> Just wan't to let you know that we're facing the same issue here. We also >> have a quite large app, using Errai and only experiencing the issue in the >> newer Chrome version. >> No workaround except disabling the optimisation yet. If you found >> something or want some more info about our setup for better understanding >> I'm very happy to help. >> >> >> Am Freitag, 22. Januar 2016 18:18:01 UTC+1 schrieb Joshb: >>> >>> Gilberto, >>> >>> I've done some more digging. Starting Chrome with >>> --js-flags="--stack-size 60000" makes the application work. I haven't >>> played with the stack size to see what level breaks it. >>> >>> Interestingly, I've tried smaller modules, and the problem persists. We >>> are using Errai, but the reason I think the issue is more to do with GWT is >>> that >>> >>> >>> - it works in SDM >>> - It works when optimizations are set to 0. In fact, the output is >>> over 70MB and it still works. >>> - it fails when optimizations are >0 >>> >>> This makes me think that some optimization being done by the GWT >>> compiler is generating a stack too deep for Chrome (works fine on IE, FF, >>> and Safari, as well as Chrome pre 47). >>> >>> My understanding of one of the promises of GWT is that it compiles >>> separate version for different browsers. I wouldn't think that code that >>> otherwise works could be broken in the compilation process. >>> >>> On Sunday, January 17, 2016 at 11:58:52 AM UTC-5, Gilberto Torrezan >>> Filho wrote: >>>> >>>> Hi, I have a large GWT app deployed too, and didn't notice any problem >>>> with the new Chrome. Do you use any large GWT framework, such as Vaadin or >>>> GXT? Or any framework/lib outside GWT on the client side? Or maybe any >>>> native component? >>>> >>>> On Wednesday, December 16, 2015 at 3:11:10 PM UTC-2, Joshb wrote: >>>>> >>>>> Hi all. >>>>> >>>>> We have a fairly large GWT application (~20MB OBF). Recently (last >>>>> week I believe) we started to encounter a RangeError: Maximum call stack >>>>> size exceed in Chrome in our deployed application. The error does not >>>>> occur in FF or IE. I downloaded older versions of Chromium and confirmed >>>>> that the application still works, so I'm guessing there was a recent >>>>> update >>>>> to Chrome (47) that caused the issue. >>>>> >>>>> There is no problem in SDM, and after playing around with the >>>>> optimization level, found that at level '0', the app loads. At 1-9, the >>>>> exception is thrown, and from what I can tell by setting a breakpoint on >>>>> exception, it does not appear to be too deep in a call stack. >>>>> >>>>> I realize that this is not a bug per-se of GWT, but I would think >>>>> (hope) that the GWT compiler could account for browser limitations when >>>>> performing optimizations, to not optimize otherwise working code into >>>>> code >>>>> that won't run. >>>>> >>>>> Would love to supply more information if needed. I'm hoping others >>>>> have discovered the same issue, and that there is an easy solution that >>>>> can >>>>> squeeze into the 2.8 release. >>>>> >>>>> Thanks in advance, >>>>> >>>>> Josh >>>>> >>>> -- >> You received this message because you are subscribed to the Google Groups >> "GWT Contributors" group. >> To unsubscribe from this group and stop receiving emails from it, send an >> email to [email protected] >> <javascript:>. >> To view this discussion on the web visit >> https://groups.google.com/d/msgid/google-web-toolkit-contributors/b996f740-4b03-4e65-ab54-b1182154256a%40googlegroups.com >> >> <https://groups.google.com/d/msgid/google-web-toolkit-contributors/b996f740-4b03-4e65-ab54-b1182154256a%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> For more options, visit https://groups.google.com/d/optout. >> > -- You received this message because you are subscribed to the Google Groups "GWT Contributors" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/google-web-toolkit-contributors/67542136-babd-4036-b72b-106f67d46c3f%40googlegroups.com. For more options, visit https://groups.google.com/d/optout.
