Re: COPTS for Makefile LinuxPPC 2000

Re: COPTS for Makefile LinuxPPC 2000

[Derek Homeier]

On Fri, 20 Oct 2000 15:17:26 -0700 (PDT), Michael Norton <mnorton@cisco.com> wrote:
>
> I am trying to build the OpenGL Heretic game and there are compiler options in t
> he Makefile for m68k but not for the PPC.
>
> COPT.m68k = -m68020-60 -O2 -fomit-frame-pointer -D__BIG_ENDIAN__ -D__32BIT__ \
>         -DHAVE_ALLOCA_H
>
>
> was should this statement read if I am going to compile a PPC image on my G4?
>
Well, you probably should keep the defines, and possibly the "-fomit-frame-pointer"
as well, so leave everything after that unchanged. It's possible you need
"-fsigned-char" in addition. You obviously wouldn't want the "-m68020-60"
which turns on m68k-specific tuning. You might simply try "-O2" or "-O3" for a
first run. There also are some architecture-specific flags available on PPC,
such as"-mcpu=604", don't know if there is a flag for 750 or even 7400 yet.
You should look at the gcc info page for more details on the different
optimizations (under "Invoking GCC::"). Chances are you won't need
terribly agressive optimization if that game actually runs on a 68k :-).

HTH,
							Derek


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: COPTS for Makefile LinuxPPC 2000

[William Blew]

On Mon, 23 Oct 2000, Derek Homeier wrote:

> such as"-mcpu=604", don't know if there is a flag for 750 or even 7400 yet.

As of gcc 2.95.2 (at least) there is a -mcpu=750 (or -mtune=750) option
and on my beige G3/266 it was good for about a 11% throughput improvement
on the utah-glx's memory bandwidth benchmarks.

--
William Blew, wblew@home.com
Gamer by Choice, Geek by Birth


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: gpm and XF401 (Re: ADB-Mouse)

[Derek Homeier]

On Mon, 23 Oct 2000 15:00:02 +0200, Olaf Hering <olh@suse.de> wrote:
>
> You mean you mark some text at console with gpm and all you get in the
> selected area is garbage, even with a non psaux device?
> I tried it here on a chrp machine and got the same effect, could be a
> framebuffer bug. Anyone else with that problem?
>
Yes, with every 2.3/2.4pre kernel I can think of (though I haven't booted a
2.4 in a while). This is on a Lombard/Rage II with atyfb.

Cheers,
							Derek


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

gpm and XF401 (Re: ADB-Mouse)

[Henning Lvser]

Hi,
thanks I got my mouse working again ... well at least sort of, I do
echo "1"  > /proc/sys/dev/mac_hid/keyboard_sends_linux_keycodes   and
echo "1"  > /proc/sys/dev/mac_hid/mouse_button_emulation.
Then the keyboard works with the .../i386/qwertz/de-latin-nodeadkeys.map.gz and even displays the 'Umlaute' in Emacs :-).
The mouse moves now but the buttons only work sort of. If I mark something in the console to copy it, it scrambles the underlying text (intead of highlithing it) and only the scrambeled text gets copied. For example it makes "n~3" from "not" and ASCII
code 'hex 02' from blanks ... Under X the mouse buttons behave funny aswell, instead of moving a window it looks as if it is resized, it's just that the part that is moved doesn't get drawn any more. If I do a xrefresh it turns out that the window was
indeed moved ... This is propably a more X related problem, I have compiled the X401 from the sources with Ani Joshies xfree86 tree and Franz's xf4-kbdppc4.patch.

Any sufggestions?

Henning

> Date: Fri, 20 Oct 2000 12:18:26 +0200
> From: Olaf Hering <olh@suse.de>
> Subject: Re: ADB-Mouse
>
> On Fri, Oct 20, Topi Kanerva wrote:
>
> > guys, how exactly do you make it to work?
>
> Put that in boot.local:
>
> echo "1"  > /proc/sys/dev/mac_hid/mouse_button_emulation
> echo "87" > /proc/sys/dev/mac_hid/mouse_button3_keycode # F11
> echo "88" > /proc/sys/dev/mac_hid/mouse_button2_keycode # F12
>
> Gruss Olaf

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: gpm and XF401 (Re: ADB-Mouse)

[Olaf Hering]

On Mon, Oct 23, Henning Löser wrote:

> Hi,
> thanks I got my mouse working again ... well at least sort of, I do
> echo "1"  > /proc/sys/dev/mac_hid/keyboard_sends_linux_keycodes   and
> echo "1"  > /proc/sys/dev/mac_hid/mouse_button_emulation.
> Then the keyboard works with the .../i386/qwertz/de-latin-nodeadkeys.map.gz and even displays the 'Umlaute' in Emacs :-).
> The mouse moves now but the buttons only work sort of. If I mark something in the console to copy it, it scrambles the underlying text (intead of highlithing it) and only the scrambeled text gets copied. For example it makes "n~3" from "not" and ASCII
> code 'hex 02' from blanks ... Under X the mouse buttons behave funny aswell, instead of moving a window it looks as if it is resized, it's just that the part that is moved doesn't get drawn any more. If I do a xrefresh it turns out that the window was
> indeed moved ... This is propably a more X related problem, I have compiled the X401 from the sources with Ani Joshies xfree86 tree and Franz's xf4-kbdppc4.patch.
>
> Any sufggestions?

You mean you mark some text at console with gpm and all you get in the
selected area is garbage, even with a non psaux device?
I tried it here on a chrp machine and got the same effect, could be a
framebuffer bug. Anyone else with that problem?

I have XF4 here with kernel 2.4 and no problems at all (fbdev driver).



Gruss Olaf

--
 $ man clone

BUGS
       Main feature not yet implemented...

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: gpm and XF401 (Re: ADB-Mouse)

[Henning Lvser]

Olaf Hering wrote:
> You mean you mark some text at console with gpm and all you get in the
> selected area is garbage, even with a non psaux device?

Right, thats it.

> I tried it here on a chrp machine and got the same effect, could be a
> framebuffer bug. Anyone else with that problem?
>
> I have XF4 here with kernel 2.4 and no problems at all (fbdev driver).
Well I use the tdfx_drv on my voodoo3 ... guess I forgot to mention that one :-)

Henning

** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: gpm and XF401 (Re: ADB-Mouse)

[William Blew]

On Mon, 23 Oct 2000, Olaf Hering wrote:

> You mean you mark some text at console with gpm and all you get in the
> selected area is garbage, even with a non psaux device?
> I tried it here on a chrp machine and got the same effect, could be a
> framebuffer bug. Anyone else with that problem?

Perhaps the hardware cursor? I have been experiencing hashed text problems
under the hardware cursor. I am using a fairly new pmac-stable 2.2. kernel
with my beige G3's builtin 3D rage pro (atyfb.c)

Specifically after an exit of the latest SDL library's console driver, the
text under the cursor is hashed, as is the width of the cursor (about 6-8
chars, not 1 char).

A "reset" command on the console fixes things, but nonetheless, it seems
to be similar, superficially.
--
William Blew, wblew@home.com
Gamer by Choice, Geek by Birth


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

Re: gpm and XF401 (Re: ADB-Mouse)

[Josh Huber]

[-- Attachment #1: Type: text/plain, Size: 790 bytes --]

On Mon, Oct 23, 2000 at 03:00:02PM +0200, Olaf Hering wrote:
> You mean you mark some text at console with gpm and all you get in the
> selected area is garbage, even with a non psaux device?
> I tried it here on a chrp machine and got the same effect, could be a
> framebuffer bug. Anyone else with that problem?
> 
> I have XF4 here with kernel 2.4 and no problems at all (fbdev driver).

I have the same problem on a g4 (r128 fb driver) when using 2.4.

I haven't had the time to look into it, but selecting text in console
mode is definately broken.  It looks like it's somehow using the wrong
characters for the highlighted versions of the text.

Anyone have any ideas,

-- 
Josh
| huber@mclx.com |
1024D/6B21489A 61F0 6138 BE7B FEBF A223  E9D1 BFE1 2065 6B21 489A

[-- Attachment #2: Type: application/pgp-signature, Size: 0 bytes --]

Re: gpm and XF401 (Re: ADB-Mouse)

[Geert Uytterhoeven]

On Mon, 23 Oct 2000, Josh Huber wrote:
> On Mon, Oct 23, 2000 at 03:00:02PM +0200, Olaf Hering wrote:
> > You mean you mark some text at console with gpm and all you get in the
> > selected area is garbage, even with a non psaux device?
> > I tried it here on a chrp machine and got the same effect, could be a
> > framebuffer bug. Anyone else with that problem?
> >
> > I have XF4 here with kernel 2.4 and no problems at all (fbdev driver).
>
> I have the same problem on a g4 (r128 fb driver) when using 2.4.
>
> I haven't had the time to look into it, but selecting text in console
> mode is definately broken.  It looks like it's somehow using the wrong
> characters for the highlighted versions of the text.
>
> Anyone have any ideas,

Perhaps it's exchanging the attribute byte with the character byte? Check out
the scr_*() macros/functions.

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
							    -- Linus Torvalds


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/

COPTS for Makefile LinuxPPC 2000

[Michael Norton]

hi,

I am trying to build the OpenGL Heretic game and there are compiler options in t
he Makefile for m68k but not for the PPC.

COPT.m68k = -m68020-60 -O2 -fomit-frame-pointer -D__BIG_ENDIAN__ -D__32BIT__ \
        -DHAVE_ALLOCA_H


was should this statement read if I am going to compile a PPC image on my G4?

thanks,

Mike


** Sent via the linuxppc-dev mail list. See http://lists.linuxppc.org/