ANDROID: mm, kasan: fix for "integrate page_alloc init with HW_TAGS"

My commit "integrate page_alloc init with HW_TAGS" changed the order of
kernel_unpoison_pages() and kernel_init_free_pages() calls. This leads
to complaints from the page unpoisoning code, as the poison pattern gets
overwritten for __GFP_ZERO allocations.

Fix by restoring the initial order. Also add a warning comment.

Reported-by: Vlastimil Babka <vbabka@suse.cz>
Reported-by: Sergei Trofimovich <slyfox@gentoo.org>
Signed-off-by: Andrey Konovalov <andreyknvl@google.com>
Reviewed-by: Sergei Trofimovich <slyfox@gentoo.org>

The fix was included in upstream patch 1bb5eab30d68 'kasan, mm: integrate page_alloc init with HW_TAGS'

Fixes: a15989497e ("FROMGIT: kasan, mm: integrate page_alloc init with HW_TAGS")
Bug: 192912968
Signed-off-by: Suren Baghdasaryan <surenb@google.com>
Change-Id: Ifc4ab7ed27b3c060f70ba5ca03117a36c9b4da08
This commit is contained in:
Andrey Konovalov
2021-03-30 18:37:36 +02:00
committed by Suren Baghdasaryan
parent f215850ea0
commit 1dd167be9f

View File

@@ -2359,6 +2359,13 @@ inline void post_alloc_hook(struct page *page, unsigned int order,
arch_alloc_page(page, order);
debug_pagealloc_map_pages(page, 1 << order);
/*
* Page unpoisoning must happen before memory initialization.
* Otherwise, the poison pattern will be overwritten for __GFP_ZERO
* allocations and the page unpoisoning code will complain.
*/
kernel_unpoison_pages(page, 1 << order);
/*
* As memory initialization might be integrated into KASAN,
* kasan_alloc_pages and kernel_init_free_pages must be
@@ -2375,7 +2382,6 @@ inline void post_alloc_hook(struct page *page, unsigned int order,
gfp_flags & __GFP_ZEROTAGS);
}
kernel_unpoison_pages(page, 1 << order);
set_page_owner(page, order, gfp_flags);
}