On Mon, 27 Nov 2017, Tejun Heo wrote:
> Hello,
>
> On Mon, Nov 27, 2017 at 03:31:52PM -0500, Nicolas Pitre wrote:
> > So IMHO I don't think reverting the commit is the right thing to do.
> > That commit is clearly not at fault here.
>
> It's not about the blame. We just want to avoid breaking boot in a
> way which is difficult to debug. Once cris is fixed, we can re-apply
> the patch.
In that case I suggest the following instead. No point penalizing
everyone for a single architecture's fault. And this will serve as a
visible reminder to the cris people that they need to clean up.
----- >8
Subject: percpu: hack to let the CRIS architecture to boot until they clean up
Commit 438a506180 ("percpu: don't forget to free the temporary struct
pcpu_alloc_info") uncovered a problem on the CRIS architecture where
the bootmem allocator is initialized with virtual addresses. Given it
has:
#define __va(x) ((void *)((unsigned long)(x) | 0x80000000))
then things just work out because the end result is the same whether you
give this a physical or a virtual address.
Untill you call memblock_free_early(__pa(address)) that is, because
values from __pa() don't match with the virtual addresses stuffed in the
bootmem allocator anymore.
Avoid freeing the temporary pcpu_alloc_info memory on that architecture
until they fix things up to let the kernel boot like it did before.
Signed-off-by: Nicolas Pitre <[email protected]>
diff --git a/mm/percpu.c b/mm/percpu.c
index 79e3549cab..50e7fdf840 100644
--- a/mm/percpu.c
+++ b/mm/percpu.c
@@ -2719,7 +2719,11 @@ void __init setup_per_cpu_areas(void)
if (pcpu_setup_first_chunk(ai, fc) < 0)
panic("Failed to initialize percpu areas.");
+#ifdef CONFIG_CRIS
+#warning "the CRIS architecture has physical and virtual addresses confused"
+#else
pcpu_free_alloc_info(ai);
+#endif
}
#endif /* CONFIG_SMP */
From 1585254113220582065@xxx Mon Nov 27 20:55:32 +0000 2017
X-GM-THRID: 1584466607656225114
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread
Hello, Nicolas.
On Mon, Nov 27, 2017 at 03:51:04PM -0500, Nicolas Pitre wrote:
> Subject: percpu: hack to let the CRIS architecture to boot until they clean up
>
> Commit 438a506180 ("percpu: don't forget to free the temporary struct
> pcpu_alloc_info") uncovered a problem on the CRIS architecture where
> the bootmem allocator is initialized with virtual addresses. Given it
> has:
>
> #define __va(x) ((void *)((unsigned long)(x) | 0x80000000))
>
> then things just work out because the end result is the same whether you
> give this a physical or a virtual address.
>
> Untill you call memblock_free_early(__pa(address)) that is, because
> values from __pa() don't match with the virtual addresses stuffed in the
> bootmem allocator anymore.
>
> Avoid freeing the temporary pcpu_alloc_info memory on that architecture
> until they fix things up to let the kernel boot like it did before.
>
> Signed-off-by: Nicolas Pitre <[email protected]>
This totally works for me. Replaced the revert with this one.
Thanks!
--
tejun
From 1585252845127743430@xxx Mon Nov 27 20:35:23 +0000 2017
X-GM-THRID: 1584466607656225114
X-Gmail-Labels: Inbox,Category Forums,HistoricalUnread