Re: 5-STABLE kernel build with icc broken
On Sat, 19 Mar 2005 13:06:29 +0100 Attila Nagy <[EMAIL PROTECTED]> wrote: > It seems to me that building kernel with icc is currently broken, at > least in 5-STABLE. Could somebody investigate this? I don't have a problem to compile it with a recent -current and a recent icc (-stable not tested), but the resulting kernel imediatly panics (page fault in _mtx_...()). Bye, Alexander. -- It's not a bug, it's tradition! http://www.Leidinger.net Alexander @ Leidinger.net GPG fingerprint = C518 BC70 E67F 143F BE91 3365 79E2 9C60 B006 3FE7 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
On Sun, Mar 27, 2005 at 01:30:59PM +0200, Alexander Leidinger wrote: > > It seems to me that building kernel with icc is currently broken, at > > least in 5-STABLE. Could somebody investigate this? > > I don't have a problem to compile it with a recent -current and a recent > icc (-stable not tested), but the resulting kernel imediatly panics > (page fault in _mtx_...()). Without intending to start any compiler holy wars, what benefits does ICC provide over GCC for the end user? ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
> >Without intending to start any compiler holy wars, what benefits does >ICC provide over GCC for the end user? > ICC would provide better low level code (remind: Intel C Compiler. It would mean better performance). rookie ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Sun, 27 Mar 2005 [EMAIL PROTECTED] wrote: Without intending to start any compiler holy wars, what benefits does ICC provide over GCC for the end user? ICC would provide better low level code (remind: Intel C Compiler. It would mean better performance). rookie If any, still produces not all that much of a difference of code between the newer gcc34 and as much performance differance as your going to get isnt going to even be noticeable in the long run. Your just setting your self up for failure with something that isnt really going to give you the desired effects. - -- Best regards, --c0ldbyte -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (FreeBSD) Comment: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xF7DF979F iD8DBQFCRwjgsmFQuvffl58RAoqKAJ44D4TFVVaHgK2bP7rrKV0cLHBGlQCeJauB ajI0mxvPps7e/l9dU14DMMU= =73/q -END PGP SIGNATURE- ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
-BEGIN PGP SIGNED MESSAGE- Hash: SHA1 On Sun, 27 Mar 2005, c0ldbyte wrote: On Sun, 27 Mar 2005 [EMAIL PROTECTED] wrote: Without intending to start any compiler holy wars, what benefits does ICC provide over GCC for the end user? ICC would provide better low level code (remind: Intel C Compiler. It would mean better performance). rookie If any, still produces not all that much of a difference of code between the newer gcc34 and as much performance differance as your going to get isnt going to even be noticeable in the long run. Your just setting your self up for failure with something that isnt really going to give you the desired effects. -- Best regards, --c0ldbyte PS: There is coders from Intel that do work on some of the code for gcc34. -BEGIN PGP SIGNATURE- Version: GnuPG v1.4.0 (FreeBSD) Comment: http://pgp.mit.edu:11371/pks/lookup?op=get&search=0xF7DF979F iD8DBQFCRwlhsmFQuvffl58RAq83AJsGKYklfVtdxeT8UcIcJ21TaqAmiQCfY6Fz JhQgmTHP66gd6ySeo0zueHc= =RrMC -END PGP SIGNATURE- ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
c0ldbyte wrote: If any, still produces not all that much of a difference of code between the newer gcc34 and as much performance differance as your going to get isnt going to even be noticeable in the long run. Your just setting your self up for failure with something that isnt really going to give you the desired effects. You don't have to use it, but it is good if you *can*. I guess besides that having a cleaner code base which compiles not only with exactly one compiler it is always good to have the ability to try something else out. BTW, on my humble Pentium II server I noticed significant speedups, compared to the system compiler. But that's purely empirical. -- Attila Nagy e-mail: [EMAIL PROTECTED] Free Software Network (FSN.HU) phone @work: +361 371 3536 ISOs: http://www.fsn.hu/?f=downloadcell.: +3630 306 6758 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
c0ldbyte wrote: PS: There is coders from Intel that do work on some of the code for gcc34. Wow. As far as I know, there are some coders from Nominum who do (or did) work on bind9. And? Bind9 is at least 10 times slower on FreeBSD than Nominum's CNS. :( I didn't get your point. -- Attila Nagy e-mail: [EMAIL PROTECTED] Free Software Network (FSN.HU) phone @work: +361 371 3536 ISOs: http://www.fsn.hu/?f=downloadcell.: +3630 306 6758 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
On Sun, Mar 27, 2005, c0ldbyte wrote: > -BEGIN PGP SIGNED MESSAGE- > Hash: SHA1 > > On Sun, 27 Mar 2005 [EMAIL PROTECTED] wrote: > > >> > >>Without intending to start any compiler holy wars, what benefits does > >>ICC provide over GCC for the end user? > >> > > > >ICC would provide better low level code (remind: Intel C Compiler. It would > >mean better performance). > > > >rookie > > > > If any, still produces not all that much of a difference of code between > the newer gcc34 and as much performance differance as your going to get > isnt going to even be noticeable in the long run. Your just setting your > self up for failure with something that isnt really going to give you > the desired effects. For some applications, particularly in scientific computing, icc is significantly better. The FreeBSD kernel is not in this category, however. Operating system kernels tend to spend most of their time chasing pointers and copying data, and compilers can't really optimize these operations. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
On Sun, 27 Mar 2005 05:40:44 -0800 Avleen Vig <[EMAIL PROTECTED]> wrote: > On Sun, Mar 27, 2005 at 01:30:59PM +0200, Alexander Leidinger wrote: > > > It seems to me that building kernel with icc is currently broken, at > > > least in 5-STABLE. Could somebody investigate this? > > > > I don't have a problem to compile it with a recent -current and a recent > > icc (-stable not tested), but the resulting kernel imediatly panics > > (page fault in _mtx_...()). > > Without intending to start any compiler holy wars, what benefits does > ICC provide over GCC for the end user? Various: - auto-vectorizer (no benefit for the kernel, since we can't use FPU/SIMD instructions at any time... yet (interested hackers can have a look how DragonFly handles it, I can provide the necessary commit logs)) - optimizations for Intel CPUs direct from the manufacturer of the CPU (they have a lot of interest to produce very fast code) - a different set of compiler warnings - better code quality (if is compilable by more than one compiler it may be more portable) Icc already pointed out some bad code (asm code in the IP checksumming code... DragonFly changed it already), and the panic as noticed above may also be an indication that we have some code in the tree which smells bad. Bye, Alexander. -- The dark ages were caused by the Y1K problem. http://www.Leidinger.net Alexander @ Leidinger.net GPG fingerprint = C518 BC70 E67F 143F BE91 3365 79E2 9C60 B006 3FE7 ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
Alexander Leidinger wrote: On Sun, 27 Mar 2005 05:40:44 -0800 Avleen Vig <[EMAIL PROTECTED]> wrote: On Sun, Mar 27, 2005 at 01:30:59PM +0200, Alexander Leidinger wrote: It seems to me that building kernel with icc is currently broken, at least in 5-STABLE. Could somebody investigate this? I don't have a problem to compile it with a recent -current and a recent icc (-stable not tested), but the resulting kernel imediatly panics (page fault in _mtx_...()). Without intending to start any compiler holy wars, what benefits does ICC provide over GCC for the end user? Various: - auto-vectorizer (no benefit for the kernel, since we can't use FPU/SIMD instructions at any time... yet (interested hackers can have a look how DragonFly handles it, I can provide the necessary commit logs)) Are you implying DragonFly uses FPU/SIMD? For that matter does any kernel? Thanks, jason - optimizations for Intel CPUs direct from the manufacturer of the CPU (they have a lot of interest to produce very fast code) - a different set of compiler warnings - better code quality (if is compilable by more than one compiler it may be more portable) Icc already pointed out some bad code (asm code in the IP checksumming code... DragonFly changed it already), and the panic as noticed above may also be an indication that we have some code in the tree which smells bad. Bye, Alexander. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"
Re: 5-STABLE kernel build with icc broken
jason henson <[EMAIL PROTECTED]> wrote: >Various: > - auto-vectorizer (no benefit for the kernel, since we can't use > FPU/SIMD instructions at any time... yet (interested hackers can > have a look how DragonFly handles it, I can provide the necessary > commit logs)) > > Are you implying DragonFly uses FPU/SIMD? For that matter does any kernel? AFAIK DragonFly _allows_ code to use the FPU/SIMD in the kernel. And AFAIR they use SIMD in b{copy,zero} (we do this too, but we do this is in a "controlled environment" whereas DFly just allows the use of FPU/SIMD in an "use as you like" manner everywhere). Bye, Alexander. -- http://www.Leidinger.net Alexander @ Leidinger.net: PGP ID = B0063FE7 http://www.FreeBSD.org netchild @ FreeBSD.org : PGP ID = 72077137 Closet extrovert. ___ freebsd-hackers@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-hackers To unsubscribe, send any mail to "[EMAIL PROTECTED]"