Re: About gcc builtin atomics

2012-06-26 Thread Thorsten Glaser
Mikael Pettersson dixit: >- having a vsyscall page at a fixed location (vdso is too horrible), and >- updating a long word in it in each switch_to() sounds ok to me >- SMP is initially prohibited, the get_thread_area entry needs to be a > proper function, so that a future SMP-capable system can

Re: About gcc builtin atomics

2012-06-26 Thread Mikael Pettersson
On Tue, 26 Jun 2012 12:29:49 + (UTC), Thorsten Glaser wrote: > >OTOH, if CAS is provably unsafe in user-space for a sizeable portion > >of the Linux/m68k HW base, then the kernel should IMO provide a > >vsyscall page with at least CAS and DCAS/CAS64 entry points (doing > > And TLS! > > So w

Re: About gcc builtin atomics

2012-06-26 Thread Geert Uytterhoeven
On Tue, Jun 26, 2012 at 2:29 PM, Thorsten Glaser wrote: >>> > (On Linux, but not on e.g. FreeMiNT.) >>> >>> FreeMiNT on which hardware? > > Doesn’t matter: it doesn’t have a Linux-specific syscall. I meant: on which hardware does FreeMiNT use CAS? Gr{oetje,eeting}s,                         Geer

Re: About gcc builtin atomics

2012-06-26 Thread Thorsten Glaser
Mikael Pettersson dixit: >> > (On Linux, but not on e.g. FreeMiNT.) >> >> FreeMiNT on which hardware? Doesn’t matter: it doesn’t have a Linux-specific syscall. >OTOH, if CAS is provably unsafe in user-space for a sizeable portion >of the Linux/m68k HW base, then the kernel should IMO provide a

Re: About gcc builtin atomics

2012-06-26 Thread Geert Uytterhoeven
On Tue, Jun 26, 2012 at 11:30 AM, Mikael Pettersson wrote: > On Tue, 26 Jun 2012 10:48:02 +0200, Geert Uytterhoeven > wrote: >> On Tue, Jun 26, 2012 at 10:26 AM, Thorsten Glaser wrote: >> > Geert Uytterhoeven dixit: >> > >> >>Even if they're fixed on ARAnyM, this doesn't fix the CAS problems on

Re: About gcc builtin atomics

2012-06-26 Thread Mikael Pettersson
On Tue, 26 Jun 2012 10:48:02 +0200, Geert Uytterhoeven wrote: > On Tue, Jun 26, 2012 at 10:26 AM, Thorsten Glaser wrote: > > Geert Uytterhoeven dixit: > > > >>Even if they're fixed on ARAnyM, this doesn't fix the CAS problems on sev= > eral > >>instantations of real hardware > > > > Right. T

Re: About gcc builtin atomics

2012-06-26 Thread Geert Uytterhoeven
On Tue, Jun 26, 2012 at 10:26 AM, Thorsten Glaser wrote: > Geert Uytterhoeven dixit: > >>Even if they're fixed on ARAnyM, this doesn't fix the CAS problems on several >>instantations of real hardware > > Right. That’s why we use the syscall always. OK. > (On Linux, but not on e.g. FreeMiNT.)

Re: About gcc builtin atomics

2012-06-26 Thread Mikael Pettersson
On Mon, 25 Jun 2012 22:14:03 + (UTC), Thorsten Glaser wrote: > Mikael Pettersson dixit: > > >Now we know that aranym had a broken > >CAS emulation :-( > > Those problems are gone now though? I hope so. I've updated my previous sync implementations for m68k (both the linux-atomic.c one and

Re: About gcc builtin atomics

2012-06-26 Thread Thorsten Glaser
Mikael Pettersson dixit: >Many things are indeed failing in the test suite on m68k. What you >should do is to make two full bootstraps and regression test runs, >first with a baseline version (I use FSF vanilla, you might want to >use Debian's common gcc without your patch), and then with the >ba

Re: About gcc builtin atomics

2012-06-26 Thread Thorsten Glaser
Geert Uytterhoeven dixit: >Even if they're fixed on ARAnyM, this doesn't fix the CAS problems on several >instantations of real hardware Right. That’s why we use the syscall always. OK. (On Linux, but not on e.g. FreeMiNT.) bye, //mirabilos -- [...] if maybe ext3fs wasn't a better pick, or

Re: About gcc builtin atomics

2012-06-26 Thread Geert Uytterhoeven
On Tue, Jun 26, 2012 at 12:14 AM, Thorsten Glaser wrote: >>Now we know that aranym had a broken >>CAS emulation :-( > > Those problems are gone now though? Even if they're fixed on ARAnyM, this doesn't fix the CAS problems on several instantations of real hardware Gr{oetje,eeting}s,