On 2/3/20 7:28 AM, Björn Töpel wrote:
On Sun, 2 Feb 2020 at 14:37, Alex Ghiti <alex@xxxxxxxx> wrote:
[...]
I think it would be better to completely avoid this patch and the
definition of this
new zone by using the generic implementation if we had the patch
discussed here
regarding modules memory allocation (that in any case we need to fix
modules loading):
https://lore.kernel.org/linux-riscv/d868acf5-7242-93dc-0051-f97e64dc4387@xxxxxxxx/T/#m2be30cb71dc9aa834a50d346961acee26158a238
This patch is already upstream. I agree that when the module
allocation fix is upstream, the BPF image allocation can be folded
into the module allocation. IOW, I wont send any page table dumper
patch for BPF memory.
Too late then :) I'll remove this zone with the patch regarding module
allocation.
But keep in mind that the RV BPF JIT relies on having the kernel text
within the 32b range (as does modules)
Yep, same constraints as for modules ;)
Thanks,
Alex
Cheers,
Björn
