+1 to what thakis@ wrote. -- Dirk
On Fri, Apr 24, 2020 at 10:29 AM K. Moon <km...@chromium.org> wrote: > Sounds like other than maybe checking again if third parties using > PDFium/V8 can switch to using clang-cl, the conclusion is to stick with the > status quo. > > On Thu, Apr 23, 2020 at 6:56 PM Nico Weber <tha...@chromium.org> wrote: > >> They're there to check that the !is_clang bits of the _gn_ build files >> mostly work, since those bits are used by non-chromium projects. >> >> On Thu, Apr 23, 2020 at 8:41 PM Adenilson Cavalcanti < >> cavalcan...@chromium.org> wrote: >> >>> As they are still available through manual selection, I would like to >>> ask what is the purpose of the MSVC bots in Chromium's infrastructure? >>> >>> It seems they don't actually build the code and execute the steps up >>> to 'analyze': >>> https://ci.chromium.org/p/webrtc/builders/try/win_x64_msvc_rel/19172? >>> >>> On Thu, Apr 23, 2020 at 3:25 PM Nico Weber <tha...@chromium.org> wrote: >>> > >>> > My thought is that we should explicitly not support this in Chromium. >>> We've talked before about how if you want chroimum's zlib to be some >>> general zlib fork, then src.git isn't the ideal place to host it. It also >>> isn't the worst place, but it comes with tradeoffs, and this is one of them. >>> > >>> > IMHO clang-cl is very stable by now, and MSVC ships with clang-cl as >>> of MSVC 2019 so we should reevaluate if we need to support msvc in pdfium, >>> v8, etc. >>> > >>> > On Thu, Apr 23, 2020 at 6:08 PM Lei Zhang <thes...@chromium.org> >>> wrote: >>> >> >>> >> We recently hit 2 MSVC issues in zlib. Aside from that, most of the >>> prior PDFium MSVC issues have been with V8. IIRC, V8's MVSC bot builds >>> 64-bit release, whereas PDFium bots build 32-bit and 64-bit, in debug mode, >>> so they do not catch the same set of issues. >>> >> >>> >> Here are all the V8 MSVC fixes from the last 2 years for issues we >>> found by rolling V8 in PDFium: >>> >> >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1964391 >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1960906 >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1960905 >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1899547 >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1697043 >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1329694 >>> >> https://chromium-review.googlesource.com/c/v8/v8/+/1187702 >>> >> >>> >> >>> >> On Thu, Apr 23, 2020 at 2:40 PM Takuto Ikuta <tik...@chromium.org> >>> wrote: >>> >>> >>> >>> If you see such issues frequently, I think you can file an issue to >>> set up the bot to detect that. >>> >>> But I currently think fixing after CQ failure in your repository >>> seems reasonable. >>> >>> >>> >>> +Nico Weber >>> >>> Do you have any thoughts for this? >>> >>> >>> >>> On Fri, Apr 24, 2020 at 6:30 AM K. Moon <km...@chromium.org> wrote: >>> >>>> >>> >>>> Hi infra folks, >>> >>>> >>> >>>> Chromium doesn't support MSVC anymore, but PDFium and V8 explicitly >>> do, and take some dependencies from Chromium, like //third_party/zlib. >>> >>>> >>> >>>> I recently ran into an issue with MSVC compilation while rolling >>> DEPS for PDFium (see crrev.com/c/2161475 for more details). The >>> breaking change to //third_party/zlib wasn't caught by the MSVC bots. (Some >>> speculation as to why on >>> https://pdfium-review.googlesource.com/c/pdfium/+/69002.) This required >>> a fix to third_party/zlib/BUILD.gn before the PDFium roll succeeded. V8 >>> likely would have had the same issue when trying to roll zlib. >>> >>>> >>> >>>> As long as PDFium and V8 continue to maintain MSVC support, I was >>> wondering if there's some way we could make sure the Chromium bots catch >>> such things? Right now, they don't seem to catch changes to zlib. >>> >>>> >>> >>>> --Kahmy >>> >>>> >>> >>>> -- >>> >>>> You received this message because you are subscribed to the Google >>> Groups "infra-dev" group. >>> >>>> To unsubscribe from this group and stop receiving emails from it, >>> send an email to infra-dev+unsubscr...@chromium.org. >>> >>>> To view this discussion on the web visit >>> https://groups.google.com/a/chromium.org/d/msgid/infra-dev/CACwGi-5gsTqbHB4WnPHWHTb_7LKkemwwfzBXe5KUFvG95GRkiA%40mail.gmail.com >>> . >>> > >>> > -- >>> > You received this message because you are subscribed to the Google >>> Groups "infra-dev" group. >>> > To unsubscribe from this group and stop receiving emails from it, send >>> an email to infra-dev+unsubscr...@chromium.org. >>> > To view this discussion on the web visit >>> https://groups.google.com/a/chromium.org/d/msgid/infra-dev/CAMGbLiEofzvqdmoDpo0uC%3DwJrLUQdNVKu425f3Acxsb_LWa8Ug%40mail.gmail.com >>> . >>> >> -- > You received this message because you are subscribed to the Google Groups > "infra-dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to infra-dev+unsubscr...@chromium.org. > To view this discussion on the web visit > https://groups.google.com/a/chromium.org/d/msgid/infra-dev/CACwGi-525cTt5NR-x1weA8rhTi9-6dJPtt5YsOgh_JW%3DN8MpMA%40mail.gmail.com > <https://groups.google.com/a/chromium.org/d/msgid/infra-dev/CACwGi-525cTt5NR-x1weA8rhTi9-6dJPtt5YsOgh_JW%3DN8MpMA%40mail.gmail.com?utm_medium=email&utm_source=footer> > . > -- -- v8-dev mailing list v8-dev@googlegroups.com http://groups.google.com/group/v8-dev --- You received this message because you are subscribed to the Google Groups "v8-dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to v8-dev+unsubscr...@googlegroups.com. To view this discussion on the web visit https://groups.google.com/d/msgid/v8-dev/CAEoffTCj2wSM_fDgSb_05Kwb5-hZZ%3DOHeEVSwLd3hYHAj1wXQw%40mail.gmail.com.
