| Message ID | 497a1d88a93.7082e0e9@auth.smtp.1and1.co.uk (mailing list archive) |
|---|---|
| State | Accepted |
| Headers | show |
On Mon, 2017-01-23 at 19:42:54 UTC, Darren Stevens wrote: > prom_init.c calls 'instance-to-package' twice, but the return > is not checked during prom_find_boot_cpu(). The result is then > passed to prom_getprop, which could be PROM_ERROR. > Add a return check to prevent this. > > This was found on a pasemi system, where CFE doesn't have a working > 'instance-to package' prom call. > Before Commit 5c0484e25ec0 ('powerpc: Endian safe trampoline') the > area around addr 0 as mostly 0's and this doesn't cause a problem. > Once the macro 'FIXUP_ENDIAN' has been added to head_64.S, the low > memory area now has non-zero values, which cause the prom_getprop > call to hang. > > Signed-off-by: Darren Stevens <darren@stevens-zone.net> Applied to powerpc fixes, thanks. https://git.kernel.org/powerpc/c/af2b7fa17eb92e52b65f96604448ff cheers
Hello Michael On 27/01/2017, Michael Ellerman wrote: > On Mon, 2017-01-23 at 19:42:54 UTC, Darren Stevens wrote: > Applied to powerpc fixes, thanks. > > https://git.kernel.org/powerpc/c/af2b7fa17eb92e52b65f96604448ff Thanks, I was going to ask about backporting to 4.9 as it's a LTS now, but I see from the above you've beaten me to it! Much appreciated. I'm nearing the point where I will want to add the Nemo setup code to the kernel, and I see 2 choices: Continue as we do, with #ifdefs in the setup.c / pci.c files. Create 2 new files setup-nemo.c and pci-nemo.c - I am assuming this would be the prefered option? Regards Darren
diff --git a/arch/powerpc/kernel/prom_init.c b/arch/powerpc/kernel/prom_init.c index ec47a93..ac83eb0 100644 --- a/arch/powerpc/kernel/prom_init.c +++ b/arch/powerpc/kernel/prom_init.c @@ -2834,6 +2834,9 @@ static void __init prom_find_boot_cpu(void) cpu_pkg = call_prom("instance-to-package", 1, 1, prom_cpu); + if (!PHANDLE_VALID(cpu_pkg)) + return; + prom_getprop(cpu_pkg, "reg", &rval, sizeof(rval)); prom.cpu = be32_to_cpu(rval);
prom_init.c calls 'instance-to-package' twice, but the return is not checked during prom_find_boot_cpu(). The result is then passed to prom_getprop, which could be PROM_ERROR. Add a return check to prevent this. This was found on a pasemi system, where CFE doesn't have a working 'instance-to package' prom call. Before Commit 5c0484e25ec0 ('powerpc: Endian safe trampoline') the area around addr 0 as mostly 0's and this doesn't cause a problem. Once the macro 'FIXUP_ENDIAN' has been added to head_64.S, the low memory area now has non-zero values, which cause the prom_getprop call to hang. Signed-off-by: Darren Stevens <darren@stevens-zone.net> ---