Re: [arch-general] [arch-dev-public] i686 and SSE2
On Sunday, 18 September 2016 07:53:56 CEST Doug Newgard wrote: > This thread is split between arch-general and arch-dev-public Ohh, my bad. Did not properly read the headers. Thank You. -- Cheers Jayesh Badwaik signature.asc Description: This is a digitally signed message part.
Re: [arch-general] [arch-dev-public] i686 and SSE2
On Sun, 18 Sep 2016 14:50:09 +0200 Jayesh Badwaikwrote: > On Saturday, 17 September 2016 15:10:48 CEST LoneVVolf wrote: > > On Fri, 2016-09-16 at 21:44 +0200, Bartłomiej Piotrowski wrote: > > > Actually, why don't raise the bar higher? SSE2 has been introduced in > > > 2001 – that's 15 years to upgrade one's hardware and given my sad > > > experiences with computers, I find it hard to believe anyone has that > > > old PC that happens to run Arch. > > > > > I did not receive the above Bartlomiej's mail from the Arch Mailing List or > even the thread starter. Was it a private mail? I need to know because > otherwise it is getting rejected by my mail provider and I need to know what? > This thread is split between arch-general and arch-dev-public
Re: [arch-general] [arch-dev-public] i686 and SSE2
On Saturday, 17 September 2016 15:10:48 CEST LoneVVolf wrote: > On Fri, 2016-09-16 at 21:44 +0200, Bartłomiej Piotrowski wrote: > > Actually, why don't raise the bar higher? SSE2 has been introduced in > > 2001 – that's 15 years to upgrade one's hardware and given my sad > > experiences with computers, I find it hard to believe anyone has that > > old PC that happens to run Arch. > > I did not receive the above Bartlomiej's mail from the Arch Mailing List or even the thread starter. Was it a private mail? I need to know because otherwise it is getting rejected by my mail provider and I need to know what? -- Cheers Jayesh Badwaik signature.asc Description: This is a digitally signed message part.
Re: [arch-general] [arch-dev-public] i686 and SSE2
I'm not sure enforcing sse3 or later for AMD64 hardware is the best course of action. > egrep "^model name|^flags" /proc/cpuinfo model name : AMD Athlon(tm) 64 Processor 3000+ flags : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov pat pse36 clflush mmx fxsr sse sse2 syscall nx mmxext fxsr_opt lm 3dnowext 3dnow rep_good nopl eagerfpu pni lahf_lm 3dnowprefetch vmmcall This machine is still able to run Arch just fine, and has saved the day when the laptop I used day to day broke down. As others have said, AMD cpus might not have sse4 unless they are quite recent. Requiring sse2 might be a reasonable cutoff, but I'm sure there will be people still using older machines that will complain if sse2 is mandatory, as I can reasonably see late pentium 3 machines still being usable for office, email or light browsing applications. -- Mauro Santos
Re: [arch-general] [arch-dev-public] i686 and SSE2
It would probably also give more arguments/hardware in need of a retro-cpu archlinux variant/project which I'd like to see and would help out with. Which in turn might be a better alternative for my Geode as well. Just my 2¢ to this. cheers! mar77i
Re: [arch-general] [arch-dev-public] i686 and SSE2
...I'm all for making arch work well for the hardware it runs well on, and I don't think expecting SSE2 is too bleeding edge wrt hardware constraints. There are other distros that work well on cpus that aren't up to par with arch's needs. I started using gentoo on my pcengines board due to already existing performance concerns for its amd geode/i[45]86 cpu. cheers! mar77i
Re: [arch-general] [arch-dev-public] i686 and SSE2
>On Fri, 2016-09-16 at 21:44 +0200, Bartłomiej Piotrowski wrote: > > For the same reason I would not complain about requiring SSE4 > > instructions for amd64. So you recommend to drop good hardware, to pollute our environment, to waste rare earth elements and things like this? [rocketmouse@archlinux ~]$ sudo hwinfo --cpu | head -8 01: None 00.0: 10103 CPU [Created at cpu.457] Unique ID: rdCR.j8NaKXDZtZ6 Hardware Class: cpu Arch: X86-64 Vendor: "AuthenticAMD" Model: 15.107.2 "AMD Athlon(tm) X2 Dual Core Processor BE-2350" Features: fpu,vme,de,pse,tsc,msr,pae,mce,cx8,apic,sep,mtrr,pge,mca,cmov,pat,pse36,clflush,mmx,fxsr,sse,sse2,ht,syscall,nx,mmxext,fxsr_opt,rdtscp,lm,3dnowext,3dnow,rep_good,nopl,extd_apicid,eagerfpu,pni,cx16,lahf_lm,cmp_legacy,svm,extapic,cr8_legacy,3dnowprefetch,vmmcall,lbrv
Re: [arch-general] [arch-dev-public] i686 and SSE2
On Fri, 2016-09-16 at 21:44 +0200, Bartłomiej Piotrowski wrote: > Actually, why don't raise the bar higher? SSE2 has been introduced in > 2001 – that's 15 years to upgrade one's hardware and given my sad > experiences with computers, I find it hard to believe anyone has that > old PC that happens to run Arch. > > We used to advertise ourselves as optimized for modern processors. Our > "i786" really should include SSE3. For the same reason I would not > complain about requiring SSE4 instructions for amd64. > >> In the 2000's AMD and Intel followed different paths with extensions. Checking https://en.wikipedia.org/wiki/Streaming_SIMD_Extensions and its links suggest AVX was the point where the paths became the same again. AVX was designed in 2008 , but first procesoors that supported it were launched in 2011 . My 2009 AMD opteron 2378 "Barcelona" processor advertises SSE, SSE2, 3DNow & 3DNowExt . I couldn't find specifics about 3DNowExt , but chances are big that software compiled with SSE3 , SSSE3 or SSE4 will crash on pre-2011 AMD processors. LW