Re: comparing x86 and powerpc laptops
> I currently own an iBook2, and have often wondered if it would have been a > better plan to go for a laptop with support for `multimedia' activities > (e.g. Altivec in a G4), and also whether I would have a copy of gcc that > would allow me to take advantage of such hardware. Hi Ed, I actually recompiled CPU intensive application on a powerbook G4 and I found that gcc-3.2.1 have options for the altivec support. I recompiled bochs using -mcpu=7450 -mtune=7450 -maltivec -mabi=altivec and the new bochs boot a complete Debian Woody system in 1'28" instead of more the 5'. I also found that -O3 generate slower executables then -O2 (strange?) Bye, Giuseppe
XFree86 not sleeping using -ben10 on a powerbook3,5
Hi, I have a powerbook3,5 with a PowerPC g4 and radeon M9. When I close the lid, the system crashes showing something like a monitor prompt. This is happening only from when I switched to the -ben10 kernel. I did many tries, using drm-trunk server and modules, using XFree86 4.3.0 from Daniel Stone and even recompiling XFree86 from CVS. You may get the packages of XFree86 4.3.99.2 that I made at deb http://eppesuigoccas.homedns.org/~giuseppe/debian testing ppcg4 I now believe that this is a problem kernel related, but it could also be because of the G4 optimizations I used. I recompiled kernel, modules, xlibmesa-drm, xfree86, pcmcia-cs using gcc-3.2 -mpowerpc-gfxopt -mcpu=7450 -mtune=7450 -maltivec -mabi=altivec -O2 Do you have any comment on this? Do you have this kind of problems with -ben10 kernel or with this compiler or switch? Thank you for your help, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mar, 2003-04-15 alle 11:15, Benjamin Herrenschmidt ha scritto: > On Tue, 2003-04-15 at 10:27, Giuseppe Sacco wrote: > > Hi, > > I have a powerbook3,5 with a PowerPC g4 and radeon M9. > > > > When I close the lid, the system crashes showing something like a > > monitor prompt. > > What does that prompt says ? You can load the system.map along > with the kernel (sysmap= in yaboot.conf) to get symbols > instead of just numbers in there, and you can then use the "t" > command to get a backtrace. vector: 400 at pc=2f6c6960, lr=c00071d8 msr=40001032, sp=dbcd7dc0 [dbcd7d10] current=dbcd600, pid=590, comm=pmud mon> then there is a cursor that blincks very very very slowly. I did not manage to press any key. > Also make sure you have /dev/apm_bios and that APM emulation is > compiled in the kernel. I think this is good since in /var/log/Xfree86.0.log I see: (II) Open APM successful Moreover: if I close the lid while at the GDM login prompt (just after booting the machine) then the apple light over the lid goes off and a small light at the bottom of the powerbook start blinking (sleep mode is started correctly?). But then when I open the lid the screen is completely black. If I do the same think after at logging in and logging out, I get the the system monitor. Bye, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mar, 2003-04-15 alle 13:16, Benjamin Herrenschmidt ha scritto: > On Tue, 2003-04-15 at 12:32, Giuseppe Sacco wrote: > > Il mar, 2003-04-15 alle 11:15, Benjamin Herrenschmidt ha scritto: > > > On Tue, 2003-04-15 at 10:27, Giuseppe Sacco wrote: > > > > Hi, > > > > I have a powerbook3,5 with a PowerPC g4 and radeon M9. > > > > > > > > When I close the lid, the system crashes showing something like a > > > > monitor prompt. > > > > > > What does that prompt says ? You can load the system.map along > > > with the kernel (sysmap= in yaboot.conf) to get symbols > > > instead of just numbers in there, and you can then use the "t" > > > command to get a backtrace. > > > > vector: 400 at pc=2f6c6960, lr=c00071d8 > > msr=40001032, sp=dbcd7dc0 [dbcd7d10] > > current=dbcd600, pid=590, comm=pmud > > mon> Now that I specified the sysmap parameter the first line changed in vector: 400 at pc=2f6c6960, lr=c00071d8 (T disable_irq_nosync+0x98) > Do you have AGP enabled ? What other driver do you have ? some USB stuffs ? How do I check it? agpgart is loaded, but in Xfree86.0.log: AGP failed to initialise -- falling back to PCI mode. Bye, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mar, 2003-04-15 alle 13:16, Benjamin Herrenschmidt ha scritto: ... > Do you have AGP enabled ? What other driver do you have ? some USB stuffs ? more informations from XFree86.0.log: > grep -i agp /var/log/XF*log | tail -11 (--) RADEON(0): Chipset: "ATI Radeon Mobility M9 Lf (AGP)" (ChipID = 0x4c66) (II) RADEON(0): AGP Fast Write disabled by default (II) RADEON(0): [agp] Mode 0x07000201 [AGP 0x106b/0x002d; Card 0x1002/0x4c66] (II) RADEON(0): [agp] 8192 kB allocated with handle 0x (II) RADEON(0): [agp] ring handle = 0x (EE) RADEON(0): [agp] Could not map ring (WW) RADEON(0): [agp] AGP failed to initialize -- falling back to PCI mode. (WW) RADEON(0): [agp] If this is an AGP card, you may want to make sure the agpgart (II) RADEON(0): Using 8 MB AGP aperture (II) RADEON(0): Using 5 MB for AGP textures (II) RADEON(0): [drm] Initialized kernel agp heap manager, 5111808 > Looks like a driver is crashing. It seems that the function crashing is called by some drivers, but the actual implementation id architecture specific. I found the source in arch/ppc/kernel/irq.c but this is all I can do. :-( Thanks for your help, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mer, 2003-04-16 alle 03:47, Michel Dänzer ha scritto: > On Die, 2003-04-15 at 14:10, Giuseppe Sacco wrote: > > Il mar, 2003-04-15 alle 13:16, Benjamin Herrenschmidt ha scritto: > > > > > Do you have AGP enabled ? [...] > > > > How do I check it? agpgart is loaded, but in Xfree86.0.log: > > > > AGP failed to initialise -- falling back to PCI mode. > > Make sure agpgart is loaded before the radeon kernel module, and that > the latter is built from drm-trunk-module-src. Actually I load agpgart using /etc/modules, while the radeon module is built from the XFree86 4.3.99.2 source (it should be the same of trunk). I'll move away agpgart from /etc/modules and I'll add pre-install radeon modprobe agpgart in /etc/modules.conf Thanks for your help, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mer, 2003-04-16 alle 03:47, Michel Dänzer ha scritto: > On Die, 2003-04-15 at 14:10, Giuseppe Sacco wrote: > > Il mar, 2003-04-15 alle 13:16, Benjamin Herrenschmidt ha scritto: > > > > > Do you have AGP enabled ? [...] > > > > How do I check it? agpgart is loaded, but in Xfree86.0.log: > > > > AGP failed to initialise -- falling back to PCI mode. > > Make sure agpgart is loaded before the radeon kernel module, and that > the latter is built from drm-trunk-module-src. Actually I load agpgart using /etc/modules, while the radeon module is built from the XFree86 4.3.99.2 source (it should be the same of trunk). I just tried to move away agpgart from /etc/modules and put a "pre-install radeon modprobe agpgart" but this did not fixed the problem. This is what I found about AGP and DRM in /var/log/XFree86.0.log (II) RADEON(0): [drm] loaded kernel module for "radeon" driver (II) RADEON(0): [drm] created "radeon" driver at busid "PCI:0:16:0" (II) RADEON(0): [drm] added 8192 byte SAREA at 0xe3448000 (II) RADEON(0): [drm] mapped SAREA 0xe3448000 to 0x3001a000 (II) RADEON(0): [drm] framebuffer handle = 0xb800 (II) RADEON(0): [drm] added 1 reserved context for kernel (II) RADEON(0): [agp] Mode 0x07000201 [AGP 0x106b/0x002d; Card 0x1002/0x4c66] (II) RADEON(0): [agp] 8192 kB allocated with handle 0x (II) RADEON(0): [agp] ring handle = 0x (EE) RADEON(0): [agp] Could not map ring (WW) RADEON(0): [agp] AGP failed to initialize -- falling back to PCI mode. (WW) RADEON(0): [agp] If this is an AGP card, you may want to make sure the agpgart kernel module is loaded before the radeon kernel module. (II) RADEON(0): [pci] 8192 kB allocated with handle 0xe544f000 (II) RADEON(0): [pci] ring handle = 0xe544f000 (II) RADEON(0): [pci] Ring mapped at 0x32047000 (II) RADEON(0): [pci] Ring contents 0x (II) RADEON(0): [pci] ring read ptr handle = 0xe555 (II) RADEON(0): [pci] Ring read ptr mapped at 0x3001c000 (II) RADEON(0): [pci] Ring read ptr contents 0x (II) RADEON(0): [pci] vertex/indirect buffers handle = 0xe5551000 (II) RADEON(0): [pci] Vertex/indirect buffers mapped at 0x32148000 (II) RADEON(0): [pci] Vertex/indirect buffers contents 0x (II) RADEON(0): [drm] register handle = 0xb000 (II) RADEON(0): [dri] Visual configs initialized
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mer, 2003-04-16 alle 13:13, Michel Dänzer ha scritto: ... > XFree86 trunk != DRI trunk != drm-trunk-module-src > > The DRM doesn't work with the UniNorth agpgart yet without > http://www.penguinppc.org/~daenzer/DRI/drm-ioremapagp.diff . You are right, I didn't know that, but I tried the drm-trunk-module and the problem is exactly the same. Moreover, if I use drm-trunk-modules instead of the drm compiled from XFree86 4.3.99.3 then DRI cannot have GL optimization with package xlibmesa-dri. With xlibmesa-dri glxgears says around 350 FPS, using dri-trunk-modules it says 100 FPS. Bye, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mer, 2003-04-16 alle 19:03, Giuseppe Sacco ha scritto: > Il mer, 2003-04-16 alle 13:13, Michel Dänzer ha scritto: > ... > > XFree86 trunk != DRI trunk != drm-trunk-module-src > > > > The DRM doesn't work with the UniNorth agpgart yet without > > http://www.penguinppc.org/~daenzer/DRI/drm-ioremapagp.diff . > > You are right, I didn't know that, but I tried the drm-trunk-module and > the problem is exactly the same. Moreover, if I use drm-trunk-modules > instead of the drm compiled from XFree86 4.3.99.3 then DRI cannot have > GL optimization with package xlibmesa-dri. > With xlibmesa-dri glxgears says around 350 FPS, using dri-trunk-modules > it says 100 FPS. I apologize for this misunderstanding. I did not installed correctly drm-trunk-modules. Now I did it right and I notice an improvement compared to the xlibmesa-dri (XFree86) since the AGP is used correctly. (II) RADEON(0): [agp] Mode 0x07000201 [AGP 0x106b/0x002d; Card 0x1002/0x4c66] (II) RADEON(0): [agp] 8192 kB allocated with handle 0xe544b000 (II) RADEON(0): [agp] ring handle = 0x (II) RADEON(0): [agp] Ring mapped at 0x32047000 (II) RADEON(0): [agp] ring read ptr handle = 0x00101000 (II) RADEON(0): [agp] Ring read ptr mapped at 0x3001c000 (II) RADEON(0): [agp] vertex/indirect buffers handle = 0x00102000 (II) RADEON(0): [agp] Vertex/indirect buffers mapped at 0x32148000 (II) RADEON(0): [agp] AGP texture map handle = 0x00302000 (II) RADEON(0): [agp] AGP Texture map mapped at 0x32348000 (II) RADEON(0): [drm] register handle = 0xb000 (II) RADEON(0): [dri] Visual configs initialized Still I have to say that sleep mode crash with the same error. I also have to say that the FPS using drm modules is around 650 FPS and not 350 as I stated before. Bye, Giuseppe
Re: can someone make a boot image with new kernel (2.4.20) $re imac 17 1ghz
Il sab, 2003-04-19 alle 23:05, pippo poppo ha scritto: > can someone make a boot image with new kernel > (2.4.20)? > so i can install debian on my imac 17" 1ghz Did you try http://people.debian.org/~branden/ibook/ ? I installed that way a Titanium G4 867Mhz 17" without problems. Then you may get kernel and xfree and dri-modules from here deb http://eppesuigoccas.homedns.org/~giuseppe/debian testing ppcg4 The only problem is that the suspend/sleep/resume cycle will not work. Bye, Giuseppe
Re: XFree86 not sleeping using -ben10 on a powerbook3,5
Il mar, 2003-04-15 alle 14:10, Giuseppe Sacco ha scritto: > Il mar, 2003-04-15 alle 13:16, Benjamin Herrenschmidt ha scritto: > > On Tue, 2003-04-15 at 12:32, Giuseppe Sacco wrote: > > > Il mar, 2003-04-15 alle 11:15, Benjamin Herrenschmidt ha scritto: > > > > On Tue, 2003-04-15 at 10:27, Giuseppe Sacco wrote: > > > > > Hi, > > > > > I have a powerbook3,5 with a PowerPC g4 and radeon M9. > > > > > > > > > > When I close the lid, the system crashes showing something like a > > > > > monitor prompt. > > > > > > > > What does that prompt says ? You can load the system.map along > > > > with the kernel (sysmap= in yaboot.conf) to get symbols > > > > instead of just numbers in there, and you can then use the "t" > > > > command to get a backtrace. > > > > > > vector: 400 at pc=2f6c6960, lr=c00071d8 > > > msr=40001032, sp=dbcd7dc0 [dbcd7d10] > > > current=dbcd600, pid=590, comm=pmud > > > mon> > > Now that I specified the sysmap parameter the first line changed in > > vector: 400 at pc=2f6c6960, lr=c00071d8 (T disable_irq_nosync+0x98) I just recompiled using 2.4.21-rc1-ben0 from CVS and now it works. I compiled agpgart and readeon as included in the kernel (not modules) so I do not use the DRM from trunk or XFree86 4.1. Thanks for your work, Ben, Giuseppe
Re: [OT] mplayer
Il ven, 2003-05-16 alle 18:23, Stefano Zacchiroli ha scritto: > I've rebuilt mplayer from sources on a powerbook 15" 1GHz with 1Gb RAM > (obviously a Debian box :-) > > I'm unable to see "decently" DVD, I get the usual message "your system > is too slow". Hi Stefano, have a look at http://freshmeat.net/articles/view/747/ and try with this .mplayer/config: monitoraspect=320:215 vo=xv fs=yes vm=yes cache=8192 alang=it dvd-device=/dev/cdrom ao=oss mc=5 autosync=30 Bye, Giuseppe
Re: mol: emulation.S: too many positional arguments
Il sab, 2003-06-14 alle 04:41, Tom Vier ha scritto: > i've tried the latest version and debian's mol-source-package, using gcc > 2.95 and 3.3. i'm using 2.4.21 headers. was an asm macro changed between > 2.4.20 and 2.4.21? does anyone have a patch for mol to bring it up to date > for 2.4.21? I had the same problem: kernel 2.4.21-rc3 and mol-modules-source 0.9.68, but then I upgraded mol-modules-source to 0.9.69 and unpacked the tar file located in /usr/src, and the problem disappeared. Maybe, you could check once more you config. Bye, Giuseppe
Re: mol runs, but won't boot
Il dom, 2003-06-15 alle 19:54, Tom Vier ha scritto: > i've tried booting an os9 boot cd, but it just flashes the disk+question > mark icon. i'm using the version from testing. it does complain that > drivers/blk.nw is missing, now that i've checked. i assume that's the blkdev > emulator for newworld machines (like my ibook). that would explain why it > won't even boot a cd. now the question is, why is blk.nm missing? Did you reported a bug? I sent a report -- using reportbug -- more than 30 hours ago but I still wait for the acknowledge. Bye, Giuseppe
unmotivated crash
Hi, I recently upgraded to the new kernel 2.4.20-rc3-ben0 and to the new compiler gcc-3.3 on a powerbook g4 15" (Radeon). In a few days I got two crashes in exactly the same way: I had some application running and, from the wmaker menu, I ran a new rxvt. This is something I usually do, but twice this lead to crash. 'Crash' means that the screen froze, the mouse and keyboard were inusable and the machine was not accessible via network. The rxvt never came up, and the menu was still opened. Do someone, by any chance(?), experienced the same problem? Bye, Giuseppe
Re: gnome2.2 broken?
Sabato, 21 Giu 2003, alle 21:39 Europe/Rome, Grzesiek Sedek ha scritto: Hi, I wonder if anyone expierienced problems with nautulitus and evolution recently? After upgrade on thursday nothing happened, but when I've rebooted my ibook gnome spit out some errors (nautlitius cant be startet) all gnome-applets are gone, and evolution is dead as well (that's on sid) is it a problem with gtk2? I had some problem with evolution (see http://bugs.debian.org/198278) from Sid but nothing related to gnome. Bye, Giuseppe
kernel panic
Hi people, lately I got many kernel panic using 2.4.21-ben1 kernel compiled with gcc-3.3 and with gcc-2.95. The message I get at the monitor prompt is kernel bug at page_alloc.c:106! vector: 0 at pc=c0034c64, lr=c0034c64 msr=9032, sp=dc5dfde0 [dc5dfd18] current=dc5de000, pid=447, comm=XFree86 does someone has the same experience? Bye, Giuseppe
Re: kernel panic
Mercoledì, 25 Giu 2003, alle 01:33 Europe/Rome, Benjamin Herrenschmidt ha scritto: On Tue, 2003-06-24 at 16:08, Giuseppe Sacco wrote: Hi people, lately I got many kernel panic using 2.4.21-ben1 kernel compiled with gcc-3.3 and with gcc-2.95. The message I get at the monitor prompt is kernel bug at page_alloc.c:106! vector: 0 at pc=c0034c64, lr=c0034c64 msr=9032, sp=dc5dfde0 [dc5dfd18] current=dc5de000, pid=447, comm=XFree86 does someone has the same experience? That should definitely not happen. What is your setup ? (what machine, anything special you do when those happen ?) can you try getting a backtrace ? (with the "t" command in xmon, then lookup the addresses in System.map) Hi Ben, this happened starting some days ago when I updated to unstable and 2.4.21-ben1 at the same time. I then tried downgrading everythink (but not the kernel) to testing and to stable. Nothing happened. I also tried many versions of XFree86 and compiled (with different compilers and optimizations) kernel and xfree and glibc (2.3.1 and 2.3.2). I finally understood that the problem is in the kernel, in version 2.4.21-ben1. The machine I use is a PowerBook G4 running at 867Mhz. What I do in order to block the kernel is just to log in using XDM. The system hangs immediately. Or, when I have the XDM window, switch to console, login and reboot. At the xmon, when I get the backtrace I see some address. I did write beside that address, the name I found in the System.map from kernel 2.4.21-ben1 (compiled with gcc 2.95). At the end I also added the start/end addresses of every function used. c0034c64 (in __free_pages_ok) c00f3888 (in agp_generic_destroy_page) c00f2bf4 (in agp_free_memory) e3378189 e3380ed0 e3377dd4 e337df54 c004e654 (in sys_ioctl) c000613c (in ret_from_syscall_1) exception:c00 [dd5f3f50] fe93930 1005700c 1029cd84 105159d4 10510144 10132048 1004c03c 10068974 1012e144 10090f3c 0fdd606c 0fdd5efc From my System.map: c0034880 t __free_pages_ok c0034d3c t rmqueue c00f2b6c T agp_free_memory c00f2c48 T agp_allocate_memory c00f3834 t agp_generic_destroy_page c00f38b8 T agp_enable c004e588 T sys_ioctl c004e7e8 T vfs_readdir c000613c T ret_from_syscall_1 c00061f0 T ret_from_syscall_2 Thank you very much for yourk, Giuseppe
Re: 2.4.21-benh2 bug?
Il mer, 2003-07-02 alle 23:30, [EMAIL PROTECTED] ha scritto: > Hi! > > i upgrade my kernel to 2.4.21-benh2 > > All was great, but when i logout kde, or try to restart xserver (kdm) or i > try to turnoff my box (tibook 15") it give me a black screen and: > > kernel BUG at page_alloc.c:106! > vector:0 at pc = c00373c0, Ir c00373c0 > msr = 9032, sp = ef751e00 [ef751d48] > current = ef75, pid = 294, comm = XFree86 > mon> This bug is already present in 2.4.21-ben1. See http://lists.debian.org/debian-powerpc/2003/debian-powerpc-200306/msg00742.html I think that there is not yet a fix. Bye, Giuseppe
wi fi 802.11g network
Hi, I am looking for a network solution to be used with debian. I would like to connect some debian machines (2 i386 laptop, 1 powerbook) using the 802.11g wifi. The powerbook is a G4 15" that is "enabled" for airport (802.11b) but not for airport extreme (802.11g), so I decided to get three PCMCIA cards and one access point. I read some test results and I think that the orinico is on of the best solutions, but I cannot find anyone using it successfully with Linux. I also found in www.PCMall.com that the orinoco 802.11b/g PC gold card is not compatible with linux. Is there anyone with experience with 802.11g and powerbook? Thanks in advance, Giuseppe
kernel panic with latest 2.4.22-ben1 from CVS
I just rsync'd from CVS the latest kernel and got a kernel panic during boot. The last line printed is Serial driver version enabled the I get the internal monitor. The backtrace is c031c14d c02dde2c c02ca4d4 c000400c c0008bed symbol map: c0008bec __main c0003ff8 init c02ca4ac do_initcalls c02dd870 rs_init c031b5e0 pidhash The machine is a powerbook3,5 [EMAIL PROTECTED], 15" with radeon, 512Mb RAM and internal airport. Does anyone have the same problem? Bye, Giuseppe
Re: kernel panic with latest 2.4.22-ben1 from CVS
Il dom, 2003-08-31 alle 10:12, Pander ha scritto: > Same here on PowerBook G4 12" > > Didn't change any settings in .config The problem seems to be solved in 2.4.22-ben2. The new kernel boots without problem, but then it panics when closing the lid (in the pmud task). Do you experience the same problem? Bye, Giuseppe
new kernel panic
... still panic ... I was syncing JPilot with an external PalmOS device, via USB, when I got this kernel panic: vector: 300 at pc = e30088b4, lr = e30088a4 msr = 9032, sp = dfe77ee0 [dfe77e30] dar = 99c, dsisr = 4200 current = dfe76000, pid = 8, comm = kbubd mon> t backtrace: e30088a4 c018de98 (c018ddec usb_disconnect) c0190d44 (c0190acc usb_hub_port_connecct_change) c019101c (c0190d48 usb_hub_events) c01910f0 (c01910b4 usb_hub_thread) c0008be0 (c0008bb4 arch_kernel_thread) kernel is 2.4.22-ben2, the machine is a powerbook3,5 15" Should I post this kind of problems on a different list? Maybe a kernel list? Bye, Giuseppe
Re: kernel panic with latest 2.4.22-ben1 from CVS
> On Thu, 2003-09-04 at 18:57, Giuseppe Sacco wrote: >> Il dom, 2003-08-31 alle 10:12, Pander ha scritto: >> > Same here on PowerBook G4 12" >> > >> > Didn't change any settings in .config >> >> The problem seems to be solved in 2.4.22-ben2. The new kernel boots >> without problem, but then it panics when closing the lid (in the pmud >> task). >> >> Do you experience the same problem? > > What panic ? Have an oops report or something ? Do you have patches on > top of bare-bones 2.4.22-ben2 ? http://lists.debian.org/debian-powerpc/2003/debian-powerpc-200308/msg00820.html no patches at all. Bye, Giuseppe Sacco
Re: debian on ibook g4??
Il dom, 2003-11-30 alle 12:23, fabrizio messina ha scritto: > hi all, > > some people just installed debian gnu linux on ibook g4? Yes, some of them are completely supported, other are less supported. What model do you have? > the question is: do you think this machine will work faster with > linux? :-) > I have not so much time time to install linux, and i would like to > spend my time > in the right way ;-) I think it isn't a matter of speed. It is a matter of freedom and usability. It is up to you to select the OS you like. Bye, Giuseppe
Re: installing from source
Il giorno mar, 25-01-2005 alle 08:26 -0500, O. T. Suarez ha scritto: > Hi: > I want to know where can I set up system-wide options if I intend to > install packages from source (and where do I find a list of available > options). Things like setting up the processor type and (possible) > optimizations. [...] Please have a look at the apt-build package. -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
powerpc and cdrecord with kernel 2.6
Hi all, I have a powerbook G4 where I get all sort of errors using cdrecord. Sometimes i do not get any error but the burned CD isn't readable. I have similar problems even on i386 and I inquired the cdrecord author. He said that there are some problems in kernels from, at least, 2.6.8. Since 2.6.8 is the default kernel for debian-powerpc, does anyone here have a working cdrecord? what kernel/program version do you use? Thanks, Giuseppe
Re: powerpc and cdrecord with kernel 2.6
Il giorno mar, 16-11-2004 alle 00:38 +0100, Sven Luther ha scritto: > On Mon, Nov 15, 2004 at 08:52:33PM +0100, Giuseppe Sacco wrote: [...] > > Sometimes i do not get any error but the burned CD isn't readable. > > > > I have similar problems even on i386 and I inquired the cdrecord author. > > He said that there are some problems in kernels from, at least, 2.6.8. > > Fixed in latest debian kernels normally, if you still see it, please fill bug > reports. I am using 2.6.9 since I am testing a patch I wish to be included in actual kernel. The patch works for 2.6.8 and 2.6.9, and it is required in order to make work a SCSI pcmcia I use. I just reported a bug, providing the patch: #281434. If this patch would be included in the debian shipped kernel, then I will gladly switch to it as soos as the kernel hit unstable. Thanks, Giuseppe
Re: jdk for ppc
Il giorno sab, 27-11-2004 alle 18:20 -0800, Andrea Giusto ha scritto: > On Saturday 27 November 2004 01:56 am, Sven Luther wrote: [...] > Ok, sorry for being so stupid but I can't figure it out by myself :( I > downloaded IBM-Java2-1.4.2-PowerPC. If I get in j2re and run ./java -version > it seems to work...but without any installation I mean: what should I do > with the downloaded directory, where linux expects it to be? I installed it in /opt/IBMJava2-ppc-142. You may install it anywhere and then you only have to adjust your PATH variable. Bye, Giuseppe
Re: Some problems in the daily use of a Debian System
Il giorno mer, 01-12-2004 alle 19:37 +0100, Pablo Gil ha scritto: > 2.6.8. I've been told that this a really bad kernel for cd writing as > well. Is it true? I tried many times, but I never managed to burn a data cd using 2.6.8, while I made once a audio CD and it worked. Bye, Giuseppe
Re: Cloning to an external video projector...
Il giorno gio, 06-01-2005 alle 12:30 +0100, Arnaud Vandyck ha scritto: > Hi, > > I'd have to make a presentation tomorrow (yes!) and I'd like to make it > on my Debian PowerBook ;-) The presentation is about a training in Linux > and the whole session (6 month, full time) will be on Debian boxes! [...] non need to patch xfree86. I have a normal sarge installation. This is the releveant part of xf86config-4: Section "Device" Identifier "radeon9000" Driver "radeon" Option "UseFBDev" "true" Option "DDCMode" "true" # Option "MonitorLayout" "LVDS,TMDS" Option "CloneMode" "1280x1024" Option "CRT2HSync" "31.5-80" Option "CRT2VRefresh" "56.3-75" EndSection man radeon for more informations. Bye, Giuseppe
cannot compile latest 2.4-benh kernel
Hi all, today I rsync'd from linux-2.4-benh and I tried to compile the kernel. I get some errors that I cannot fix. Could anyone help me? gcc -D__KERNEL__ -I/usr/src/linux-2.4-benh/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -I/usr/src/linux-2.4-benh/arch/ppc -fsigned-char -msoft-float -pipe -ffixed-r2 -Wno-uninitialized -mmultiple -mstring -nostdinc -iwithprefix include -DKBUILD_BASENAME=nvram -c -o nvram.o nvram.c nvram.c: In function `nvram_ioctl': nvram.c:78: error: `OBSOLETE_PMAC_NVRAM_GET_OFFSET' undeclared (first use in this function) nvram.c:78: error: (Each undeclared identifier is reported only once nvram.c:78: error: for each function it appears in.) nvram.c:79: error: `IOC_NVRAM_GET_OFFSET' undeclared (first use in this function) nvram.c:90: error: `IOC_NVRAM_SYNC' undeclared (first use in this function) nvram.c:91: warning: implicit declaration of function `pmac_nvram_sync' make[4]: *** [nvram.o] Error 1 make[4]: Leaving directory `/usr/src/linux-2.4-benh/drivers/macintosh' make[3]: *** [first_rule] Error 2 Thanks, Giusepe
Re: cannot compile latest 2.4-benh kernel
Il mar, 2003-12-16 alle 19:45, Joe Malik ha scritto: > On Tue, Dec 16, 2003 at 03:05:42PM +0100, Giuseppe Sacco wrote: > > Hi all, > > today I rsync'd from linux-2.4-benh and I tried to compile the kernel. I > > get some errors that I cannot fix. Could anyone help me? > > > > gcc -D__KERNEL__ -I/usr/src/linux-2.4-benh/include -Wall > > -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common > > -fomit-frame-pointer -I/usr/src/linux-2.4-benh/arch/ppc -fsigned-char > > -msoft-float -pipe -ffixed-r2 -Wno-uninitialized -mmultiple -mstring > > -nostdinc -iwithprefix include -DKBUILD_BASENAME=nvram -c -o nvram.o > > nvram.c > > nvram.c: In function `nvram_ioctl': > > nvram.c:78: error: `OBSOLETE_PMAC_NVRAM_GET_OFFSET' undeclared (first > > use in this function) [...] > IIRC I ran into the same problem as a result of combing nvram, generic- > rtc and extended-rtc in an invalid way. Though I don't remember what > combination exactly caused the failure; look for NVRAM and RTC in the > config file and/or (s)diff against my config to find the relevant entries. > http://www.daniel-zimmermann.com/joe/ibook/config-2.4.23-pre5-ben0joe9 Thanks joe, I did try as you suggested but It did not fix the problem. Moreover I think that something is broken in the kernel. Look at this eppesuig3:/usr/src/linux-2.4-benh# grep -r IOC_NVRAM_GET_OFFSET . ./drivers/macintosh/nvram.c:case IOC_NVRAM_GET_OFFSET: { eppesuig3:/usr/src/linux-2.4-benh# Anyway thanks, Giuseppe
new cdrom burner on powerbook doesn't work
Hi all, I just replaced the internal cdrom of a powerbook itanium g4. The previous one stopped working some months ago when the laptop fallen down :-( The new one (CD/DVD reader, CD writer) is the same model as the previous one (8121A) and it works quite well sith MacOSX. With my Sarge I may read CD and DVD but I cannot burn new CD. I have two different behaviours: 1. sometimes the new media is just ejected 2. sometimes the new media is "accepted", then I run cdrecord and it recognize the media too. then cdrecord tell me that the driver need to reload the media (this is different than what happened with the previous drive, but cdrecord is also a new version), it ejects the media, I reinsert the media and the burning starts. cdrecord complete the burning operation without errors but then the cdrom isn't readable, i.e., it cannot be mounted (even on a different computer). Does anyone have an idea on what to do? Thanks, Giuseppe
pcmcia with 2.4.32-ben1
Hi all, I have a SCSI pcmcia card that needs the sym53c500_cs. Actually I have PCMCIA support built in the kernel and alls PCMCIA SCSI driver built as modules, but when I insert the card I see, in syslog: cardmgr[503]: socket 0: New Media Bus Toaster SCSI cardmgr[503]: executing: 'modprobe sym53c500_cs' cardmgr[503]: + modprobe: Can't locate module sym53c500_cs cardmgr[503]: modprobe exited with status 255 cardmgr[503]: module /lib/modules/2.4.23-ben1/pcmcia/sym53c500_cs.o not available cardmgr[503]: get dev info on socket 0 failed: Resource temporarily unavailable So I am trying to compile the pcmcia-cs package. I had to disable PCMCIA support in the kernel, but I cannot compile the package I got from unstable because of this error: cc -fno-builtin -msoft-float -ffixed-r2 -MD -O3 -Wall -Wstrict-prototypes -pipe -Wa,--no-warn -I../include -I/usr/src/linux-2.4-benh/include -D__KERNEL__ -DMODULE -c cs.c In file included from ../include/asm/io.h:5, from /usr/src/linux-2.4-benh/include/asm/dma.h:12, from /usr/src/linux-2.4-benh/include/asm/scatterlist.h:5, from /usr/src/linux-2.4-benh/include/asm/pci.h:8, from /usr/src/linux-2.4-benh/include/linux/pci.h:679, from ../include/linux/pci.h:4, from cs.c:49: /usr/src/linux-2.4-benh/include/asm/io.h: In function `virt_to_bus': /usr/src/linux-2.4-benh/include/asm/io.h:235: error: `CONFIG_KERNEL_START' undeclared (first use in this function) /usr/src/linux-2.4-benh/include/asm/io.h:235: error: (Each undeclared identifier is reported only once /usr/src/linux-2.4-benh/include/asm/io.h:235: error: for each function it appears in.) [...] Do you have idea of how to fix this? Thanks, Giuseppe
Re: Debian 3 disk 1 bootable?
Il mer, 2004-01-07 alle 22:14, Gaétan Quentin ha scritto: > At least i tried to download woody...with jigdo...without success. There are > 107 missing files, whichever mirror i use did you use both debian and debian-non-US sources? I do use jigdo regularly on both i386 and powerpc arches fo woody and sarge without problems. Bye, Giuseppe
Re: Kernel compilation fails
Il mar, 2004-01-20 alle 23:32, Michael Rex ha scritto: > Quoth marco <[EMAIL PROTECTED]>: > > Ok... I have the SAME problem since a month or more. > > I used to compile my kernels from there. > > I'm using MOL and I need to upgrade it... Compiling the mol-modules > > needs the source code of the kernel you will run mol under... > > Where did you get the sources from? > rsync.penguinppc.org::linux-2.4-benh? > Those from rsync.theorie.physik.uni-muenchen.de::linuxppc-2.4-benh work > for me. I did tried many times syncing from the start of the year (from penguinppc) and I always had the same error. Bye, Giuseppe
Re: Re: Build errors for 2.4.25-ben1
In-Reply-To=<[EMAIL PROTECTED]> Hi Wolfgang, today rsync from rsync.penguinppc.org::linux-2.4-benh give me to the same error you had some days ago: [...] cp vmlinux debian/tmp-image/boot/vmlinux-2.4.25-ben1 chmod 644 debian/tmp-image/boot/vmlinux-2.4.25-ben1 cp arch/ppc/boot/images/vmlinux.coff debian/tmp-image/boot/vmlinux.coff-2.4.25-ben1 cp: cannot stat `arch/ppc/boot/images/vmlinux.coff': No such file or directory make[1]: *** [real_stamp_image] Error 1 make[1]: Leaving directory `/usr/src/linux-2.4-benh' make: *** [kernel-image-deb] Error 2 Did you find any solution? Thanks, Giuseppe
debian on apple powerbook
Hello, I just installed (again) debian on an Apple PowerBook G4. It is a quite old machine, but it fits my requirements. After having installed the latest supported Debian release, I upgraded to sid using the ports archive. A few think do not work, so I am asking how to fix them, if possible. First, kernels 5.6 and 5.7 do not boot at all. I opened a bug a few days back (https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=963689). Second, the battery is not detected by upower, so it does not show up in MATE. Third, audio is not working at all. Do you have hints about how to solve these problems? I am willing to spend some of my time in order to debug and, eventually, provide patches. Thank you, Giuseppe
Re: debian on apple powerbook
Hello Jeroen, thank you for your hints. The first about the battery worked immediately, while the second about the sound didn't. Il giorno mar, 30/06/2020 alle 08.17 +0200, Jeroen Diederen ha scritto: [...] > sound: > To get sound working install alsa-utils and launch alsamixer > > sudo apt install alsa-utils > alsamixer I have alsa-utils 1.2.2-1 from unstable already installed. # LANG=C alsamixer cannot open mixer: No such file or directory do I need any more module loaded? Thanks, Giuseppe
Re: debian on apple powerbook
Hello Jeroen, Il giorno mar, 30/06/2020 alle 09.54 +0200, Jeroen Diederen ha scritto: > For some reason, Debian does not install the correct PATH variable, > maybe that is the problem. > So you could try as as root: > > export PATH=$PATH:/usr/local/sbin:/usr/sbin:/sbin I think the PATH is already setup correctly. As root: # echo $PATH /usr/local/sbin:/usr/local/bin:/usr/sbin:/usr/bin:/sbin:/bin # alsamixer cannot open mixer: File o directory non esistente Alsa modules are not loaded, I think, since: # cat /proc/asound/cards --- no soundcards --- # cat /proc/asound/devices 33:: timer # cat /proc/asound/pcm # What is in these files in a working system? Which modules are loaded? These are mine: # lsmod | fgrep snd | sort snd57428 5 snd_aoa_i2sbus,snd_timer,snd_aoa,snd_pcm,snd_aoa_codec_tas snd_aoa13256 1 snd_aoa_codec_tas snd_aoa_codec_tas 9805 0 snd_aoa_i2sbus 18268 0 snd_aoa_soundbus3685 1 snd_aoa_i2sbus snd_pcm78991 1 snd_aoa_i2sbus snd_timer 19655 1 snd_pcm soundcore 5242 1 snd Bye, Giuseppe
Re: debian on apple powerbook
Hello Adrian, Il giorno mar, 30/06/2020 alle 10.11 +0200, John Paul Adrian Glaubitz ha scritto: > On 6/30/20 9:54 AM, Jeroen Diederen wrote: > > For some reason, Debian does not install the correct PATH variable, maybe > > that is the problem. > > So you could try as as root: > > > > export PATH=$PATH:/usr/local/sbin:/usr/sbin:/sbin > You most likely used sudo or su without using a new login shell, no? I used "su -" for going from normal user to root. I tried in two ways: remotely via ssh and on an xterm at the console. Thank you, Giuseppe
Re: debian on apple powerbook
Hello Adrian, Il giorno mar, 30/06/2020 alle 10.30 +0200, John Paul Adrian Glaubitz ha scritto: [...] > Try using Pulse-Audio. On my iBook G4, sounds works out of the box with > Pulse-Audio without any additional configuration. Your sound card is > probably just muted. This does not work either. $ pacmd | tail -4 0 card(s) available. 0 sink input(s) available. 0 source output(s) available. 0 cache entrie(s) available. > Also, software like alsa-mixer is usually run as the user logged in, > not root. Right. I have the same error with my normal user. Bye, Giuseppe
How to collect debug information while booting kernel on powerpc 32 bit
Hello, I cannot boot a powerbook g4 with kernels more recent than 3.16. Using 5.6 and 5.7 from sid, the system blocks after printing these lines: pmac32_cpufreq: registering PowerMac CPU frequency driver pamc32_cpufreq: Low: 667 MHz, High: 867 Mhz, Boot: 667 MHz I opened a bug report (#963689) but I would like to contribute more information. How may I debug the problem? I tried a lot of kernel parameters without having any additional information on the screen. Would it be helpful to connect a serial cable (where?), use it to connect a console, and get more info? Is there any parameter that may be used for collecting more information when creating the initrd image? Thank you, Giuseppe
Re: How to collect debug information while booting kernel on powerpc 32 bit
Hello Adrian, Il giorno mer, 01/07/2020 alle 20.02 +0200, John Paul Adrian Glaubitz ha scritto: > On 7/1/20 4:51 PM, Giuseppe Sacco wrote: > > How may I debug the problem? I tried a lot of kernel parameters without > > having any additional information on the screen. Would it be helpful to > > connect a serial cable (where?), use it to connect a console, and get > > more info? Is there any parameter that may be used for collecting more > > information when creating the initrd image? > > Try setting up a netconsole [1][2], that usually works just fine. Thank you for those links. I was not aware of such console, so I spent some time studying it. Unfortunately, I cannot use it since the kernel stops before listing the Ethernet device. Another option i would like to try, is to test kernel versions in between 3.16 and 5.7, in order to find about when the problem start. I couldn't find a guide for cross compiling the debian kernel package. Do you know if it is possible to cross compile the powerpc kernel on a debian 10 machine with amd64 arch? Thank you, Giuseppe
finding who call all __init functions
Hello, I am still investigating the problem about kernel not booting on powerbook. I compiled many kernel versions and found the problem starts with 5.6 since 5.4.50 boots correctly. Now, I would like to better follow the boot process, looking for what is done after the cpufreq driver setup. I cannot use the netconsole since the ethernet driver is not yet loaded when the boot process hangs. I already created a .config that include only a reduced set of kernel parts, but it is still quite large, of course. I would like to start knowing what is called after the cpufreq. How do I find it? I know that drivers/cpufreq/pmac32-cpufreq.c has an __init marked function and it is not compiled as a module. I understood that such marked function will be discarded once boot is complete, but I suspect somewhere a part of the kernel load all __init marked functions. Is this correct? Where is it? Thank you, Giuseppe
Re: finding who call all __init functions
Hello Adrian Il giorno mar, 07/07/2020 alle 12.30 +0200, John Paul Adrian Glaubitz ha scritto: > Hi Guiseppe! > On 7/7/20 11:54 AM, Giuseppe Sacco wrote [...] > > I would like to start knowing what is called after the cpufreq. How > > do I find it? I know that drivers/cpufreq/pmac32-cpufreq.c has an > > __init marked function and it is not compiled as a module. I > > understood that such marked function will be discarded once boot is > > complete, but I suspect somewhere a part of the kernel load all > > __init marked functions. Is this correct? Where is it? > > Why aren't you bisecting the kernel using git so you find the actual > commit that broke the build? This way you will find the problem much > quicker. This is a possibility, but I was unaware that the bisect works between branches (5.4 and 5.6). And, each recompile from scratch last about 6-7 hours, so I prefer to work on a single kernel version, adding printk somewhere and rebuilding only what is changed. > Also, I would suggest discussing this kernel problem on the Linux > kernel mailing list for PowerPC as the people there are actually > responsible for the code you are trying to debug [1]. Thank you, I will ask for help there now. Bye, Giuseppe
Kernel broken for Apple powerbooks
Hello, since two months I am debugging the kernel boot process on an Apple powerbook because latest Debian kernels do not boot. I already opened a bug against debian src:linux package[1]. Finally, I found the problem and I reported it upstream [2]. It seems the fix will require some time, but in the meantime, the kernel may be built and work if CONFIG_VMAP_STACK is not set. Since powerpc is not currently a released architecture, who should I bug for changing this option for the kernels available at https://www.ports.debian.org/? Or, is there any ppc32 porter able to propose a patch? Thank you, Giuseppe [1]https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=963689 [2]https://github.com/linuxppc/issues/issues/326
Re: Kernel broken for Apple powerbooks
Hello Adrian, Il giorno ven, 28/08/2020 alle 15.25 +0200, John Paul Adrian Glaubitz ha scritto: > Hi Guiseppe! > > On 8/28/20 3:14 PM, Giuseppe Sacco wrote: > > since two months I am debugging the kernel boot process on an Apple > > powerbook because latest Debian kernels do not boot. I already > > opened a bug against debian src:linux package[1]. > > Finally, I found the problem and I reported it upstream [2]. > > Thanks for tracing this down. I'm not sure, however, whether your bug > report > will get any attention there. It's better to post it on the official > kernel > mailing list called linuxppc-dev [1] and the kernel bugzilla [2]. Yes, I contacted linuxppc-dev and got much help. There I've got the request to open the bug at github. I am sure they will find a fix, but I have no ideas about how much time this will take. > > It seems the fix will require some time, but in the meantime, the > > kernel may be built and work if CONFIG_VMAP_STACK is not set > > Since powerpc is not currently a released architecture, who should I > > bug for changing this option for the kernels available at https://www.ports.debian.org/> > ? > > > > Or, is there any ppc32 porter able to propose a patch? > > You still report your bugs against src:linux. Even in Debian Ports, virtually > all source changes are made through the normal packages in unstable. We > don't have separate package sources in Debian Ports. Ok, so I will ask for this config change on the standard package src:linux using the already opened bug. Thank you, Giuseppe
Re: Kernel broken for Apple powerbooks
Hello Adrian, Il giorno ven, 28/08/2020 alle 15.51 +0200, John Paul Adrian Glaubitz ha scritto: > Hello! > > On 8/28/20 3:41 PM, Giuseppe Sacco wrote: [...] > > Ok, so I will ask for this config change on the standard package > > src:linux using the already opened bug. > > Yes, I posted a proposed patch below. You can adjust the title of > your bug > report using the BTS control server. > > diff --git a/debian/changelog b/debian/changelog > index 9d9eddea4b6a..af0a845ae5e9 100644 > --- a/debian/changelog > +++ b/debian/changelog > @@ -1,3 +1,10 @@ > +linux (5.8.3-1~exp2) UNRELEASED; urgency=medium > + > + [ John Paul Adrian Glaubitz ] > + * [powerpc] Set CONFIG_VMAP_STACK=n to unbreak kernel on Apple PowerBook. > + > + -- John Paul Adrian Glaubitz > Fri, 28 Aug 2020 15:48:32 +0200 > + > linux (5.8.3-1~exp1) experimental; urgency=medium > >* New upstream release: > https://kernelnewbies.org/Linux_5.8 > > diff --git a/debian/config/powerpc/config.powerpc > b/debian/config/powerpc/config.powerpc > index 63a07fb2e8f7..de4135891e5c 100644 > --- a/debian/config/powerpc/config.powerpc > +++ b/debian/config/powerpc/config.powerpc > @@ -1,3 +1,8 @@ > +## > +## file: arch/Kconfig > +## > +CONFIG_VMAP_STACK=n > + > ## > ## file: arch/powerpc/Kconfig > ## > > Let me try it on my on Apple iBook G4 first though. did you try on your machine? BTW, it seems this patch will be included in upstream version 5.9-rc4. Bye, Giuseppe
Re: Kernel broken for Apple powerbooks
Il giorno mar, 01/09/2020 alle 12.17 +0200, John Paul Adrian Glaubitz ha scritto: > Hi Guiseppe! > > On 9/1/20 9:30 AM, Giuseppe Sacco wrote: > > > Let me try it on my on Apple iBook G4 first though. > > > > did you try on your machine? > > Not yet. I wasn't home the past days. Will try it today. > > > BTW, it seems this patch will be included > > in upstream version 5.9-rc4. > > What patch? Did the kernel guys already fix the problem? no, the one that disable VMAP_STACK on ppc32 with PMU. And it will be included in 5.8-stable, as well. Bye, Giuseppe