Problem with 2.2.1 compilation

Problem with 2.2.1 compilation

[Marc Dejardin]

Hello,

For a long time now I encountered regularly kernel panics with recent kernels.
Every thing was incredibly stable with 2.1.24.

To make some tests I am trying to compile the kernel 2.2.1 with -O0 option
instead of -O2. Then I got the message :

make[1]: Leaving directory `/mnt/extra2/scratch/linux-pmac/arch/ppc/lib'
ld -T arch/ppc/vmlinux.lds -Ttext 0xc0000000 -Bstatic arch/ppc/kernel/head.o
init/main.o init/version.o \
        --start-group \
        arch/ppc/kernel/kernel.o arch/ppc/mm/mm.o arch/ppc/lib/lib.o
kernel/kernel.o mm/mm.o fs/fs.o ipc/ipc.o \
        fs/filesystems.a \
        net/network.a \
        drivers/block/block.a drivers/char/char.a drivers/misc/misc.a
drivers/net/net.a drivers/scsi/scsi.a drivers/cdrom/cdrom.a
drivers/sound/sound.a drivers/pci/pci.a drivers/macintosh/macintosh.a
drivers/video/video.a \
        /tmp/scratch/linux-pmac/lib/lib.a \
        --end-group \
        -o vmlinux
init/main.o: In function `calibrate_delay':
init/main.o(.text.init+0x608): undefined reference to `__delay'
init/main.o(.text.init+0x6e4): undefined reference to `__delay'
make: *** [vmlinux] Error 1

at the end ... I don't get such a message with -O2 or -O1 option.

Do I make a mistake somewhere ??? is it not possible to remove optimisation in
kernel building ?

Thanks in advance,
                                 Marc Dejardin



[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

RE: Problem with 2.2.1 compilation

[Gary Thomas]

I don't want to get into or repeat a *very lengthy* discussion here, but this
is an artifact of EGCS and inline functions.  It basically boils down to the
fact that running without -On and inline functions (the way they are used
by the Linux kernel) just doesn't jive.

Is there some reason (other than your instability which I understand) that
you want to do this?  BTW - Linux 2.2.1 is incredibly stable on my PreP box,
so I have little fear in the compiler, at least for the non-PowerMac specific
stuff.

On 12-Feb-99 Marc Dejardin wrote:
> 
> Hello,
> 
> For a long time now I encountered regularly kernel panics with recent kernels.
> Every thing was incredibly stable with 2.1.24.
> 
> To make some tests I am trying to compile the kernel 2.2.1 with -O0 option
> instead of -O2. Then I got the message :
> 
> make[1]: Leaving directory `/mnt/extra2/scratch/linux-pmac/arch/ppc/lib'
> ld -T arch/ppc/vmlinux.lds -Ttext 0xc0000000 -Bstatic arch/ppc/kernel/head.o
> init/main.o init/version.o \
>         --start-group \
>         arch/ppc/kernel/kernel.o arch/ppc/mm/mm.o arch/ppc/lib/lib.o
> kernel/kernel.o mm/mm.o fs/fs.o ipc/ipc.o \
>         fs/filesystems.a \
>         net/network.a \
>         drivers/block/block.a drivers/char/char.a drivers/misc/misc.a
> drivers/net/net.a drivers/scsi/scsi.a drivers/cdrom/cdrom.a
> drivers/sound/sound.a drivers/pci/pci.a drivers/macintosh/macintosh.a
> drivers/video/video.a \
>         /tmp/scratch/linux-pmac/lib/lib.a \
>         --end-group \
>         -o vmlinux
> init/main.o: In function `calibrate_delay':
> init/main.o(.text.init+0x608): undefined reference to `__delay'
> init/main.o(.text.init+0x6e4): undefined reference to `__delay'
> make: *** [vmlinux] Error 1
> 
> at the end ... I don't get such a message with -O2 or -O1 option.
> 
> Do I make a mistake somewhere ??? is it not possible to remove optimisation in
> kernel building ?
> 
> Thanks in advance,
>                                  Marc Dejardin
> 
> 

------------------------------------------------------------------------
Gary Thomas                              |
email: gdt@linuxppc.org                  | "Fine wine is a necessity of
   ... opinions expressed here are mine  |        life for me"
       and no one else would claim them! |
                                         |      Thomas Jefferson
------------------------------------------------------------------------



[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

RE: Problem with 2.2.1 compilation

[Marc Dejardin]

On Fri, 12 Feb 1999, Gary Thomas wrote:

> I don't want to get into or repeat a *very lengthy* discussion here, but this
> is an artifact of EGCS and inline functions.  It basically boils down to the
> fact that running without -On and inline functions (the way they are used
> by the Linux kernel) just doesn't jive.
> 
> Is there some reason (other than your instability which I understand) that
> you want to do this?  BTW - Linux 2.2.1 is incredibly stable on my PreP box,
> so I have little fear in the compiler, at least for the non-PowerMac specific
> stuff.
> 

Thanks for the answer !

My problem is that I have regural crashes (kernel panic) with recents kernels
when I run Xpmac or XF68_FBDev. (see January archive)
It seems that I am the only people in this situation on this earth, so I am
trying to solve my problem with my own and humble capacities.
Since I am not a guru neither an expert, I just play with only tuning
capapbilities I got. Compiler optimiser is one of them.

By the way, I have not crashed since I run with the 2.2.1 kernel compile with
-O1 option. Ok It is just 10 hours old but it smells better than before.

I agree that it could hide some other problems like hardware ones... but if the
problem is a timing problem with hardware, running a little bit slower could
help. In fact I am more or less convince that my problem is linked to harware
due to the fact that I have never succeeded to make the virtual memory
working correctly with MacOS (7-5-3 to 8-5-1). But if I send my old 8500/180 to
repair it will cost me the price of a new iMac... and I am even not sure that it
will work correctly after.

What bothers me is that it was working like a charm with 2.1.24 (just normal
shut down for hollidays and no crash for weeks). The point is that it seems
that X seems to work slightly faster with recent kernel ... Is it just an
illusion ?

Best regards, 
                                 Marc Dejardin


[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]

RE: Problem with 2.2.1 compilation

[Gary Thomas]

On 12-Feb-99 Marc Dejardin wrote:
> On Fri, 12 Feb 1999, Gary Thomas wrote:
> 
>> I don't want to get into or repeat a *very lengthy* discussion here, but this
>> is an artifact of EGCS and inline functions.  It basically boils down to the
>> fact that running without -On and inline functions (the way they are used
>> by the Linux kernel) just doesn't jive.
>> 
>> Is there some reason (other than your instability which I understand) that
>> you want to do this?  BTW - Linux 2.2.1 is incredibly stable on my PreP box,
>> so I have little fear in the compiler, at least for the non-PowerMac specific
>> stuff.
>> 
> 
> Thanks for the answer !
> 
> My problem is that I have regural crashes (kernel panic) with recents kernels
> when I run Xpmac or XF68_FBDev. (see January archive)
> It seems that I am the only people in this situation on this earth, so I am
> trying to solve my problem with my own and humble capacities.
> Since I am not a guru neither an expert, I just play with only tuning
> capapbilities I got. Compiler optimiser is one of them.
> 
> By the way, I have not crashed since I run with the 2.2.1 kernel compile with
> -O1 option. Ok It is just 10 hours old but it smells better than before.
> 
> I agree that it could hide some other problems like hardware ones... but if the
> problem is a timing problem with hardware, running a little bit slower could
> help. In fact I am more or less convince that my problem is linked to harware
> due to the fact that I have never succeeded to make the virtual memory
> working correctly with MacOS (7-5-3 to 8-5-1). But if I send my old 8500/180 to
> repair it will cost me the price of a new iMac... and I am even not sure that it
> will work correctly after.
> 

Sounds like your hardware may be a bit shakey...

> What bothers me is that it was working like a charm with 2.1.24 (just normal
> shut down for hollidays and no crash for weeks). The point is that it seems
> that X seems to work slightly faster with recent kernel ... Is it just an
> illusion ?
> 

Indeed the newer kernels are quite fast!  Also, 2.1.24 is pretty old, a lot
changed in ~100 revision...

Hopefully someone closer to the PowerMac side can help more.

[[ This message was sent via the linuxppc-dev mailing list. Replies are ]]
[[ not forced back to the list, so be sure to  Cc linuxppc-dev  if your ]]
[[ reply is of general interest. To unsubscribe from linuxppc-dev, send ]]
[[ the message 'unsubscribe' to linuxppc-dev-request@lists.linuxppc.org ]]