Losts of tests failed on loong64 *with* the `Unsigned`, and they would pass if I remove the `Unsigned` for loong64.
It's because LoongArch64, MIPS64 and perhaps RISCV64 all generally need to sign-extend 32-bit values in 64-bit registers according to their calling convention. But the convert to unsigned then to intptr will generate a zero-extended 32-bit value, which caused the failue on loong64 port. Not sure why x64 needs the `Unsigned`. On Wednesday, August 7, 2024 at 5:51:42 PM UTC+8 [email protected] wrote: > But will these tests not also fail on loong64 then? Why do you expect > different behavior there? > > On Wed, Aug 7, 2024 at 11:44 AM Zhao Jiazhong <[email protected]> > wrote: > >> Some tests like `wasm-spec-tests/address` failed on x64 if remove the >> `Unsigned`. >> >> On Wednesday, August 7, 2024 at 5:29:32 PM UTC+8 [email protected] wrote: >> >>> I think it should not be a problem to remove the `Unsigned` there for >>> other platforms as well, if it works. What's happening there is that a Smi >>> gets converted into an int32 and then stored in a 64-bit slot on the stack. >>> If the second half of the 64-bit slot gets filled with ones or with zeroes >>> does not matter, I think. If it matters, then there are tests in place. >>> >>> >>> On Wed, Aug 7, 2024 at 11:06 AM Zhao Jiazhong <[email protected]> >>> wrote: >>> >>>> It's a good idea, I will use this method. Thanks! >>>> >>>> On Wednesday, August 7, 2024 at 5:00:53 PM UTC+8 [email protected] >>>> wrote: >>>> >>>>> I would give this "condition" a name >>>>> `@if(NEEDS_SPECIAL_INT32_TO_INT64_SIGN_EXTENSION)` >>>>> (maybe find something more descriptive) and then set this build flag >>>>> for all architectures that need that. >>>>> >>>>> On Wed, Aug 7, 2024 at 10:52 AM Zhao Jiazhong <[email protected]> >>>>> wrote: >>>>> >>>>>> Hi Nico, >>>>>> >>>>>> Thanks for your information, it indeed works! >>>>>> But it seems that the @if and @ifnot annotations can't handle two >>>>>> conditions at once like `@if(V8_TARGET_ARCH_LOONG64 || >>>>>> V8_TARGET_ARCH_MIPS64)`, and I didn't find something like `@elif`. >>>>>> I thinks MIPS64 is likely need this change too, but add @if(MIPS64) >>>>>> and @ifnot(MIPS64) in @ifnot(LOONG64) seems ugly. Do you have any >>>>>> suggestions? Thank you very much! >>>>>> >>>>>> Yours, >>>>>> Zhao Jiazhong >>>>>> On Wednesday, August 7, 2024 at 2:21:24 PM UTC+8 [email protected] >>>>>> wrote: >>>>>> >>>>>>> Hi Zhao, >>>>>>> >>>>>>> you can use Torque's @if and @ifnot annotations to make such >>>>>>> distinctions (check @if(TAGGED_SIZE_8_BYTES) for an example). You then >>>>>>> need >>>>>>> to set this from the C++ side in torque-parser.cc >>>>>>> <https://source.chromium.org/chromium/chromium/src/+/main:v8/src/torque/torque-parser.cc;l=46;drc=35bb990bac45ef15807c9eab4d60b8078b65f038> >>>>>>> and >>>>>>> for that you can use the usual `V8_TARGET_ARCH_XXX`. The places where >>>>>>> you >>>>>>> can use such annotations are a bit restricted if I remember correctly, >>>>>>> but >>>>>>> it should be enough to support your case (maybe see this >>>>>>> <https://source.chromium.org/chromium/chromium/src/+/main:v8/src/builtins/promise-misc.tq;l=82;drc=7857eb34db42f339b337c6bdfb0d10deb14862f3> >>>>>>> >>>>>>> for an example). Hope that helps. >>>>>>> >>>>>>> Cheers >>>>>>> Nico >>>>>>> >>>>>>> >>>>>>> On Wed, Aug 7, 2024 at 4:18 AM Zhao Jiazhong <[email protected]> >>>>>>> wrote: >>>>>>> >>>>>>>> Hi all, >>>>>>>> >>>>>>>> I'm porting JSPI to loong64 port, and find an issue that in >>>>>>>> JSToWasmWrapperHelper >>>>>>>> <https://source.chromium.org/chromium/chromium/src/+/refs/tags/119.0.6045.214:v8/src/builtins/js-to-wasm.tq;l=559>, >>>>>>>> >>>>>>>> a 32-bit value is converted to unsigned then converted to intptr, >>>>>>>> which >>>>>>>> leading to a zero-extended value, but on LoongArch64, we need the >>>>>>>> 32-bit >>>>>>>> value to be sign-extended in 64-bit registers. >>>>>>>> >>>>>>>> I don't want to change the behavior on other arches, but the >>>>>>>> builtin is written in torque, I suppose I can't use >>>>>>>> `V8_TARGET_ARCH_XXX` >>>>>>>> macro in it, so is there a way to distinguish v8 target arch in >>>>>>>> torque builtins? Thanks! >>>>>>>> >>>>>>>> Yours, >>>>>>>> Zhao Jiazhong >>>>>>>> >>>>>>>> -- >>>>>>>> -- >>>>>>>> v8-dev mailing list >>>>>>>> [email protected] >>>>>>>> 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 [email protected]. >>>>>>>> To view this discussion on the web visit >>>>>>>> https://groups.google.com/d/msgid/v8-dev/c2790cc7-513d-4296-8531-f620fe93e038n%40googlegroups.com >>>>>>>> >>>>>>>> <https://groups.google.com/d/msgid/v8-dev/c2790cc7-513d-4296-8531-f620fe93e038n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>>>> . >>>>>>>> >>>>>>> >>>>>>> >>>>>>> -- >>>>>>> Nico Hartmann | Software Engineer | [email protected] | Chrome - >>>>>>> V8 >>>>>>> >>>>>> -- >>>>>> -- >>>>>> v8-dev mailing list >>>>>> [email protected] >>>>>> 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 [email protected]. >>>>>> >>>>> To view this discussion on the web visit >>>>>> https://groups.google.com/d/msgid/v8-dev/88402770-d013-4847-8688-cfdf6cf0793an%40googlegroups.com >>>>>> >>>>>> <https://groups.google.com/d/msgid/v8-dev/88402770-d013-4847-8688-cfdf6cf0793an%40googlegroups.com?utm_medium=email&utm_source=footer> >>>>>> . >>>>>> >>>>> >>>>> >>>>> -- >>>>> Nico Hartmann | Software Engineer | [email protected] | Chrome - V8 >>>>> >>>> -- >>>> -- >>>> v8-dev mailing list >>>> [email protected] >>>> 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 [email protected]. >>>> >>> To view this discussion on the web visit >>>> https://groups.google.com/d/msgid/v8-dev/3e4f56ab-a3ab-4e12-941b-9319c2e29a68n%40googlegroups.com >>>> >>>> <https://groups.google.com/d/msgid/v8-dev/3e4f56ab-a3ab-4e12-941b-9319c2e29a68n%40googlegroups.com?utm_medium=email&utm_source=footer> >>>> . >>>> >>> >>> >>> -- >>> >>> Andreas Haas >>> >>> Software Engineer >>> >>> [email protected] >>> >>> >>> Google Germany GmbH >>> >>> Erika-Mann-Straße 33 >>> >>> 80636 München >>> >>> >>> Geschäftsführer: Paul Manicle, Liana Sebastian >>> >>> Registergericht und -nummer: Hamburg, HRB 86891 >>> >>> Sitz der Gesellschaft: Hamburg >>> >>> >>> Diese E-Mail ist vertraulich. Falls Sie diese fälschlicherweise erhalten >>> haben sollten, leiten Sie diese bitte nicht an jemand anderes weiter, >>> löschen Sie alle Kopien und Anhänge davon und lassen Sie mich bitte wissen, >>> dass die E-Mail an die falsche Person gesendet wurde. >>> >>> >>> >>> This e-mail is confidential. If you received this communication by >>> mistake, please don't forward it to anyone else, please erase all copies >>> and attachments, and please let me know that it has gone to the wrong >>> person. >>> >>> -- >> -- >> v8-dev mailing list >> [email protected] >> 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 [email protected]. >> > To view this discussion on the web visit >> https://groups.google.com/d/msgid/v8-dev/06ce8d10-ee9a-4993-946e-1c6bc3e352cdn%40googlegroups.com >> >> <https://groups.google.com/d/msgid/v8-dev/06ce8d10-ee9a-4993-946e-1c6bc3e352cdn%40googlegroups.com?utm_medium=email&utm_source=footer> >> . >> > > > -- > > Andreas Haas > > Software Engineer > > [email protected] > > > Google Germany GmbH > > Erika-Mann-Straße 33 > > 80636 München > > > Geschäftsführer: Paul Manicle, Liana Sebastian > > Registergericht und -nummer: Hamburg, HRB 86891 > > Sitz der Gesellschaft: Hamburg > > > Diese E-Mail ist vertraulich. Falls Sie diese fälschlicherweise erhalten > haben sollten, leiten Sie diese bitte nicht an jemand anderes weiter, > löschen Sie alle Kopien und Anhänge davon und lassen Sie mich bitte wissen, > dass die E-Mail an die falsche Person gesendet wurde. > > > > This e-mail is confidential. If you received this communication by > mistake, please don't forward it to anyone else, please erase all copies > and attachments, and please let me know that it has gone to the wrong > person. > > -- -- v8-dev mailing list [email protected] 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 [email protected]. To view this discussion on the web visit https://groups.google.com/d/msgid/v8-dev/6d148e25-963e-4c60-b7b5-5f7c11c64e2fn%40googlegroups.com.
