Alberto Barbaro via llvm-dev
2019-Dec-18 07:44 UTC
[llvm-dev] Standard way to remove llvm intrinsics
Hi Eric, Currently I'm mainly playing with interpreter and trying to understand how it exactly works. I have noticed that lli complains when llvm.dbg.value is encounter that why I wanted to remove it. Additionally, I would like to understand that performances with and without intrinsics and the impact they have on the interpreter in general. Thanks Alberto On Wed, Dec 18, 2019, 00:33 Eric Christopher <echristo at gmail.com> wrote:> Typically this isn't done, debug information is a bit of a special case. I > don't see much of a reason to have tool support for this necessarily, but > would be happy to hear about possible use cases etc? > > Thanks! > > -eric > > On Mon, Dec 16, 2019 at 12:04 PM Alberto Barbaro via llvm-dev < > llvm-dev at lists.llvm.org> wrote: > >> Hi all, >> it is not the first time that I need to remove intrinsics from a bc and >> in the past I saw options like -strip-debug for opt. Now I would like to >> remove all the occurrences of llvm.lifetime.start and llvm.lifetime.end. I >> have done in just obtaining the .ll file, grep -v lifetime, redirecting it >> to another .ll and use llvm-as. >> >> I was wondering if there is a standard way to decide which intrinsics to >> remove or if at the moment it is not possible. >> >> Thanks >> Alberto >> _______________________________________________ >> LLVM Developers mailing list >> llvm-dev at lists.llvm.org >> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >> >-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20191218/6adb7360/attachment.html>
Eric Christopher via llvm-dev
2019-Dec-18 21:40 UTC
[llvm-dev] Standard way to remove llvm intrinsics
Not a lot of us use the interpreter unfortunately - probably best to just change it to ignore those intrinsics rather than error or try to remove them. -eric On Wed, Dec 18, 2019 at 12:00 AM Alberto Barbaro <barbaro.alberto at gmail.com> wrote:> Hi Eric, > Currently I'm mainly playing with interpreter and trying to understand how > it exactly works. I have noticed that lli complains when llvm.dbg.value is > encounter that why I wanted to remove it. Additionally, I would like to > understand that performances with and without intrinsics and the impact > they have on the interpreter in general. > > Thanks > Alberto > > On Wed, Dec 18, 2019, 00:33 Eric Christopher <echristo at gmail.com> wrote: > >> Typically this isn't done, debug information is a bit of a special case. >> I don't see much of a reason to have tool support for this necessarily, but >> would be happy to hear about possible use cases etc? >> >> Thanks! >> >> -eric >> >> On Mon, Dec 16, 2019 at 12:04 PM Alberto Barbaro via llvm-dev < >> llvm-dev at lists.llvm.org> wrote: >> >>> Hi all, >>> it is not the first time that I need to remove intrinsics from a bc and >>> in the past I saw options like -strip-debug for opt. Now I would like to >>> remove all the occurrences of llvm.lifetime.start and llvm.lifetime.end. I >>> have done in just obtaining the .ll file, grep -v lifetime, redirecting it >>> to another .ll and use llvm-as. >>> >>> I was wondering if there is a standard way to decide which intrinsics to >>> remove or if at the moment it is not possible. >>> >>> Thanks >>> Alberto >>> _______________________________________________ >>> LLVM Developers mailing list >>> llvm-dev at lists.llvm.org >>> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >>> >>-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20191218/85db2212/attachment.html>
Alberto Barbaro via llvm-dev
2019-Dec-18 22:09 UTC
[llvm-dev] Standard way to remove llvm intrinsics
Thanks Eric, I'll spend some time on it and try to understand how to ignore them. Alberto Il giorno mer 18 dic 2019 alle ore 21:40 Eric Christopher < echristo at gmail.com> ha scritto:> Not a lot of us use the interpreter unfortunately - probably best to just > change it to ignore those intrinsics rather than error or try to remove > them. > > -eric > > On Wed, Dec 18, 2019 at 12:00 AM Alberto Barbaro < > barbaro.alberto at gmail.com> wrote: > >> Hi Eric, >> Currently I'm mainly playing with interpreter and trying to understand >> how it exactly works. I have noticed that lli complains when llvm.dbg.value >> is encounter that why I wanted to remove it. Additionally, I would like to >> understand that performances with and without intrinsics and the impact >> they have on the interpreter in general. >> >> Thanks >> Alberto >> >> On Wed, Dec 18, 2019, 00:33 Eric Christopher <echristo at gmail.com> wrote: >> >>> Typically this isn't done, debug information is a bit of a special case. >>> I don't see much of a reason to have tool support for this necessarily, but >>> would be happy to hear about possible use cases etc? >>> >>> Thanks! >>> >>> -eric >>> >>> On Mon, Dec 16, 2019 at 12:04 PM Alberto Barbaro via llvm-dev < >>> llvm-dev at lists.llvm.org> wrote: >>> >>>> Hi all, >>>> it is not the first time that I need to remove intrinsics from a bc and >>>> in the past I saw options like -strip-debug for opt. Now I would like to >>>> remove all the occurrences of llvm.lifetime.start and llvm.lifetime.end. I >>>> have done in just obtaining the .ll file, grep -v lifetime, redirecting it >>>> to another .ll and use llvm-as. >>>> >>>> I was wondering if there is a standard way to decide which intrinsics >>>> to remove or if at the moment it is not possible. >>>> >>>> Thanks >>>> Alberto >>>> _______________________________________________ >>>> LLVM Developers mailing list >>>> llvm-dev at lists.llvm.org >>>> https://lists.llvm.org/cgi-bin/mailman/listinfo/llvm-dev >>>> >>>-------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20191218/d7b677cf/attachment.html>