On Wed Sep 19, 2007 at 09:01:38 +0000, chenggh wrote:
Looks the same as the other mail. Enter helps? Actually, what do you expect to see? :)
Thanks. I also read another mail with the same situation. I use Slackware 12.0 not Ubuntu because in Ubuntu there is a glibc bug which has been existed at least for more than half year - "undefined reference to __stack_chk_fail". This bug has been reported many times and has committed to the Ubuntu Bug List. But bug is as before.
That's no bug, it's a stack protection feature they use. I didn't really care because I do not use Ubuntu myself but I think we have to work around this.
The first question: the l4 boot log is like this: l4lx | l4lx_thread_create: Created thread 10.16 (IRQ5) l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 4. l4lx | l4lx_thread_create: Created thread 10.17 (IRQ4) l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 3. l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 2. l4lx | acquire_irq: Error attaching to IRQ 2 l4lx | l4lx_thread_create: Created thread 10.19 (IRQ2) l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 1. l4lx | acquire_irq: Error attaching to IRQ 1 It seems that when attaching IRQ1, IRQ2, IRQ9, IRQ12 there are some errors like "acquire_irq: Error attaching to IRQ x". But later attaching IRQ is ok. What is the matter? IRQ1 and IRQ12 are related with keyboard and monse, IRQ2 is related with cascade, IRQ9 is about acpi. I have removed these drivers from linux. why is it also like this?
Think about who is supposed to have these IRQs. Keyboard and Mouse are already taken by the windowing/console system, so Linux can obviously not get them. So it is really ok that attaching those IRQs fail, it's supposed to be like this.
The second problem is: The Linux booting stops at /ect/random-seed to initialize /dev/urandom and there is no login interface. I have checked the normal linux booting and after the using /ect/random-seed to initialize /dev/urandom the linux will finish booting kernel, enter the linux user space "INIT: Entering init X". So I think it is still the problem of l4linux kernel not the special linux distribution. Is it right?
Same question: Is your Slackware starting a login on tty1? What happens with normal Linux? You may want to disable your X for the test and check your /etc/inittab.
Adam
Adam Lackorzynski 写道:
On Wed Sep 19, 2007 at 09:01:38 +0000, chenggh wrote:
Looks the same as the other mail. Enter helps? Actually, what do you expect to see? :)
Thanks. I also read another mail with the same situation. I use Slackware 12.0 not Ubuntu because in Ubuntu there is a glibc bug which has been existed at least for more than half year - "undefined reference to __stack_chk_fail". This bug has been reported many times and has committed to the Ubuntu Bug List. But bug is as before.
This is not bug, but a feature. Ubuntu enables this option by default, use -fno-stack-protector in your kernel makefile to disable it. But for user space process, I suggest you enable it.
That's no bug, it's a stack protection feature they use. I didn't really care because I do not use Ubuntu myself but I think we have to work around this.
The first question: the l4 boot log is like this: l4lx | l4lx_thread_create: Created thread 10.16 (IRQ5) l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 4. l4lx | l4lx_thread_create: Created thread 10.17 (IRQ4) l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 3. l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 2. l4lx | acquire_irq: Error attaching to IRQ 2 l4lx | l4lx_thread_create: Created thread 10.19 (IRQ2) l4lx | l4lx_irq_dev_startup_hw: Starting IRQ thread for IRQ 1. l4lx | acquire_irq: Error attaching to IRQ 1 It seems that when attaching IRQ1, IRQ2, IRQ9, IRQ12 there are some errors like "acquire_irq: Error attaching to IRQ x". But later attaching IRQ is ok. What is the matter? IRQ1 and IRQ12 are related with keyboard and monse, IRQ2 is related with cascade, IRQ9 is about acpi. I have removed these drivers from linux. why is it also like this?
Think about who is supposed to have these IRQs. Keyboard and Mouse are already taken by the windowing/console system, so Linux can obviously not get them. So it is really ok that attaching those IRQs fail, it's supposed to be like this.
The second problem is: The Linux booting stops at /ect/random-seed to initialize /dev/urandom and there is no login interface. I have checked the normal linux booting and after the using /ect/random-seed to initialize /dev/urandom the linux will finish booting kernel, enter the linux user space "INIT: Entering init X". So I think it is still the problem of l4linux kernel not the special linux distribution. Is it right?
Same question: Is your Slackware starting a login on tty1? What happens with normal Linux? You may want to disable your X for the test and check your /etc/inittab.
Adam
l4-hackers@os.inf.tu-dresden.de