glibc - capaibility control mechanism

2005-04-01 Thread Bastian Blank
Situation: Current glibc have a mechanism called hwcaps. It is not documented and I was never able to correctly use it anyway. Parts of this mechanism are used to have TLS libraries. There are several problems with that: - Several hacks to use that for tls. - Not working. Proposal: * The capabil

Re: glibc - capaibility control mechanism

2005-04-01 Thread Daniel Jacobowitz
On Fri, Apr 01, 2005 at 10:12:36AM +0200, Bastian Blank wrote: > Situation: > > Current glibc have a mechanism called hwcaps. It is not documented and > I was never able to correctly use it anyway. Parts of this mechanism are > used to have TLS libraries. > There are several problems with that: >

Re: glibc - capaibility control mechanism

2005-04-01 Thread Bastian Blank
On Fri, Apr 01, 2005 at 11:35:49AM -0500, Daniel Jacobowitz wrote: > You need to be a lot more specific than that. It works. I use it every day. So? Than please show me how to ask it to enable sse optimized libs and disable tls. > No way. An additional data source with a grammar that needs to

Re: glibc - capaibility control mechanism

2005-04-01 Thread Daniel Jacobowitz
On Fri, Apr 01, 2005 at 07:11:04PM +0200, Bastian Blank wrote: > On Fri, Apr 01, 2005 at 11:35:49AM -0500, Daniel Jacobowitz wrote: > > You need to be a lot more specific than that. It works. I use it every > > day. > > So? Than please show me how to ask it to enable sse optimized libs and > di

Re: glibc - capaibility control mechanism

2005-04-01 Thread Bastian Blank
On Fri, Apr 01, 2005 at 01:27:26PM -0500, Daniel Jacobowitz wrote: > On Fri, Apr 01, 2005 at 07:11:04PM +0200, Bastian Blank wrote: > > So? Than please show me how to ask it to enable sse optimized libs and > > disable tls. > Debian doesn't offer a way to do this for libc, because we only ship > SS

Re: glibc - capaibility control mechanism

2005-04-01 Thread Daniel Jacobowitz
On Fri, Apr 01, 2005 at 08:57:56PM +0200, Bastian Blank wrote: > On Fri, Apr 01, 2005 at 01:27:26PM -0500, Daniel Jacobowitz wrote: > > On Fri, Apr 01, 2005 at 07:11:04PM +0200, Bastian Blank wrote: > > > So? Than please show me how to ask it to enable sse optimized libs and > > > disable tls. > >

Re: glibc - capaibility control mechanism

2005-04-01 Thread Bastian Blank
On Fri, Apr 01, 2005 at 03:29:06PM -0500, Daniel Jacobowitz wrote: > On Fri, Apr 01, 2005 at 08:57:56PM +0200, Bastian Blank wrote: > > On Fri, Apr 01, 2005 at 01:27:26PM -0500, Daniel Jacobowitz wrote: > > > On Fri, Apr 01, 2005 at 07:11:04PM +0200, Bastian Blank wrote: > > > > So? Than please sho

Re: glibc - capaibility control mechanism

2005-04-01 Thread Daniel Jacobowitz
On Fri, Apr 01, 2005 at 10:40:51PM +0200, Bastian Blank wrote: > On Fri, Apr 01, 2005 at 03:29:06PM -0500, Daniel Jacobowitz wrote: > > On Fri, Apr 01, 2005 at 08:57:56PM +0200, Bastian Blank wrote: > > > On Fri, Apr 01, 2005 at 01:27:26PM -0500, Daniel Jacobowitz wrote: > > > > On Fri, Apr 01, 200

Re: glibc - capaibility control mechanism

2005-04-01 Thread GOTO Masanori
At Fri, 1 Apr 2005 21:56:03 -0500, Daniel Jacobowitz wrote: > > >it will not disable SSE. > > > > SSE is masked off by default, I asked for how to enable it. > > Build a glibc which considers it a relevant capability for library > selection. It's just a matter of setting

Re: glibc - capaibility control mechanism

2005-04-02 Thread Bastian Blank
On Fri, Apr 01, 2005 at 09:56:03PM -0500, Daniel Jacobowitz wrote: > TLS isn't a hwcap. It's supported in a similar fashion, that's all. > In practice, it is available iff your kernel is new enough, so tying it > to the kernel version makes sense. It is defined as a hwcap internal. > Build a gli

Re: glibc - capaibility control mechanism

2005-04-02 Thread Bastian Blank
On Sat, Apr 02, 2005 at 03:44:22PM +0900, GOTO Masanori wrote: > However it just changes HWCAP value - so we need to make ld.so to > search library path using some library search path modifying > environment variable (ex: LD_LIBRARY_PATH). No, this will be fucking slow, thats why my proposal only

Re: glibc - capaibility control mechanism

2005-04-02 Thread GOTO Masanori
At Sat, 2 Apr 2005 10:31:32 +0200, Bastian Blank wrote: > On Sat, Apr 02, 2005 at 03:44:22PM +0900, GOTO Masanori wrote: > > However it just changes HWCAP value - so we need to make ld.so to > > search library path using some library search path modifying > > environment variable (ex: LD_LIBRARY_PA

Re: glibc - capaibility control mechanism

2005-04-03 Thread Bastian Blank
On Sat, Apr 02, 2005 at 09:37:54PM +0900, GOTO Masanori wrote: > > > I think the simple shell script wrapper controlling HWCAP_MASK can > > > achieve the original request by Bastian. > > No, it does not. You have to replace any binary with this wrapper, even > > /sbin/init. > Hmm, I have not unders

Re: glibc - capaibility control mechanism

2005-04-03 Thread GOTO Masanori
At Sun, 3 Apr 2005 12:24:38 +0200, Bastian Blank wrote: > [1 ] > On Sat, Apr 02, 2005 at 09:37:54PM +0900, GOTO Masanori wrote: > > > > I think the simple shell script wrapper controlling HWCAP_MASK can > > > > achieve the original request by Bastian. > > > No, it does not. You have to replace any

Re: glibc - capaibility control mechanism

2005-04-04 Thread Bastian Blank
On Mon, Apr 04, 2005 at 03:42:25PM +0900, GOTO Masanori wrote: > Note that 3 can be disabled on debian: /etc/ld.so.nohwcap. Tough it > disables all hwcap mechanisms, we cannot select each hwcap bits > currently. So it is unusable if there exists i386, i486 and i686. > For me, it seems this issue

Re: glibc - capaibility control mechanism

2005-04-04 Thread GOTO Masanori
At Mon, 4 Apr 2005 09:55:28 +0200, Bastian Blank wrote: > [1 ] > On Mon, Apr 04, 2005 at 03:42:25PM +0900, GOTO Masanori wrote: > > Note that 3 can be disabled on debian: /etc/ld.so.nohwcap. Tough it > > disables all hwcap mechanisms, we cannot select each hwcap bits > > currently. > > So it is