Rom Walton, from the other thread: "I've held off putting this into 7.2 for now. It turns out that detecting the CPU instruction set is only half of the problem.
We are in the process of getting an updated toolchain for building on Windows which will allow us to detect if the OS is properly saving/restoring the new registers on a thread context switch." (http://lists.ssl.berkeley.edu/pipermail/boinc_dev/2013-August/020264.html) Rom Walton, from the other thread: "We are in the process of updating our Windows toolchain to a more recent version of Visual Studio. It isn't enough to just check if the instruction set is present, we also need to check whether the OS is saving and restoring the registers as well. That required a more recent assembler than what we are currently using. Once we have the updated toolchain we will finish up with the rest of the detection and get it ported over to the 7.2 branch." (http://lists.ssl.berkeley.edu/pipermail/boinc_dev/2013-August/020316.html) On Sun, Aug 25, 2013 at 1:45 PM, Radim Vančo <rad...@centrum.cz> wrote: > I know but because it was another problem it the mailing than I suppose that > they solved this but forgot on these new CPUs. As I said, another problem so > that is why I wrote about it. > > > Dne 25.8.2013 13:41, Jord van der Elst napsal(a): > >> BOINC can not report parts of the CPU that it isn't programmed to report. >> So when for instance in 2014 there will be CPUs available with FMA3 >> and FMA4, you'll first need a new BOINC capable of detecting those >> options before that same BOINC will report them. >> Even having code added to the source code does not automatically mean >> that any of the presently available BOINC versions is capable of >> reporting that CPU option. The BOINC with the code in it will first >> have to be built & be available for download. Which in this case, it >> hasn't. >> >> >> On Sun, Aug 25, 2013 at 1:35 PM, Radim Vančo <rad...@centrum.cz> wrote: >>> >>> I hope so, because in that mailing part there was problem with not >>> reporting >>> AVX on os that doesn't support it, but it doesn't report it even on os >>> that >>> does support it. >>> >>> >>> Dne 25.8.2013 13:14, Jord van der Elst napsal(a): >>> >>>> That's because 'the latest version' doesn't support it yet. >>>> It'll be in a 7.2 version, AFTER 7.2.11, as soon as the developers >>>> port it to the 7.2 branch. >>>> >>>> If you can't wait, you can build a newer BOINC from Master. >>>> >>>> >>>> http://boinc.berkeley.edu/trac/changeset/9b2d0f992f6669d31d1996c4fc105ce22324445b/boinc-v2 >>>> shows that Rom added the change on the 22nd of July 2013 to the Master >>>> branch. >>>> It just hasn't been ported to the 7.0 or 7.2 branches yet. >>>> >>>> On Sun, Aug 25, 2013 at 12:56 PM, Radim Vančo <rad...@centrum.cz> wrote: >>>>> >>>>> I have read it, but this is another problem. The processor has AVX >>>>> support >>>>> and OS also. It is Win 7 with SP1 64bit but it doesn't report AVX. Even >>>>> with >>>>> the latest version, but it doesn't report it on these type of >>>>> processors. >>>>> i7 >>>>> reports it ok, but i5-3230M and other 3xxx doesn't report it. >>>>> >>>>> >>>>> >>>>> Dne 25.8.2013 12:01, Jord van der Elst napsal(a): >>>>> >>>>>> You probably haven't had an answer because there's just been a whole >>>>>> email thread about this. >>>>>> Can be found in the archives of the boinc_dev email list, starting >>>>>> from >>>>>> >>>>>> http://lists.ssl.berkeley.edu/pipermail/boinc_dev/2013-July/020215.html >>>>>> >>>>>> On Sun, Aug 25, 2013 at 11:49 AM, Radim Vančo <rad...@centrum.cz> >>>>>> wrote: >>>>>>> >>>>>>> No one to reply this? This is quite important, I already wrote with >>>>>>> some >>>>>>> people and they also have undetected AVX on this type of processor >>>>>>> which >>>>>>> has >>>>>>> AVX. So there is some problem in BOINC client, that it doesn't report >>>>>>> AVX. >>>>>>> >>>>>>> Radim >>>>>>> >>>>>>> >>>>>>> Dne 23.8.2013 11:51, Radim Vančo napsal(a): >>>>>>> >>>>>>>> I am now testing 7.0.64 and it also doesn't detect AVX. So it >>>>>>>> propably >>>>>>>> something in source code, that it doesn't detect on that processor. >>>>>>>> >>>>>>>> >>>>>>>> Dne 23.8.2013 11:41, Radim Vančo napsal(a): >>>>>>>>> >>>>>>>>> Does BOINC client version 7.2.10/7.2.11 detect AVX? I just realized >>>>>>>>> that >>>>>>>>> mine doesn't, but it should report. Processor i5 3230i and Win 7 >>>>>>>>> 64bit >>>>>>>>> with >>>>>>>>> SP1 installed. Isn't there a bug in BOINC, that it doesn't detect >>>>>>>>> it? >>>>>>>>> _______________________________________________ >>>>>>>>> boinc_dev mailing list >>>>>>>>> boinc_dev@ssl.berkeley.edu >>>>>>>>> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev >>>>>>>>> To unsubscribe, visit the above URL and >>>>>>>>> (near bottom of page) enter your email address. >>>>>>>> >>>>>>>> >>>>>>>> _______________________________________________ >>>>>>>> boinc_dev mailing list >>>>>>>> boinc_dev@ssl.berkeley.edu >>>>>>>> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev >>>>>>>> To unsubscribe, visit the above URL and >>>>>>>> (near bottom of page) enter your email address. >>>>>>> >>>>>>> >>>>>>> _______________________________________________ >>>>>>> boinc_dev mailing list >>>>>>> boinc_dev@ssl.berkeley.edu >>>>>>> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev >>>>>>> To unsubscribe, visit the above URL and >>>>>>> (near bottom of page) enter your email address. >>>>>> >>>>>> >>>>>> >>>>> _______________________________________________ >>>>> boinc_dev mailing list >>>>> boinc_dev@ssl.berkeley.edu >>>>> http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev >>>>> To unsubscribe, visit the above URL and >>>>> (near bottom of page) enter your email address. >>>> >>>> >>>> >> >> > -- -- Jord van der Elst. _______________________________________________ boinc_dev mailing list boinc_dev@ssl.berkeley.edu http://lists.ssl.berkeley.edu/mailman/listinfo/boinc_dev To unsubscribe, visit the above URL and (near bottom of page) enter your email address.
