All of lore.kernel.org
 help / color / mirror / Atom feed
* [folded-merged] mm-add-__dump_folio-fix-3.patch removed from -mm tree
@ 2024-03-05 20:47 Andrew Morton
  0 siblings, 0 replies; only message in thread
From: Andrew Morton @ 2024-03-05 20:47 UTC (permalink / raw)
  To: mm-commits, ryan.roberts, broonie, ardb, aishwarya.tcv, willy,
	akpm


The quilt patch titled
     Subject: mm-add-__dump_folio-fix-3
has been removed from the -mm tree.  Its filename was
     mm-add-__dump_folio-fix-3.patch

This patch was dropped because it was folded into mm-add-__dump_folio.patch

------------------------------------------------------
From: Matthew Wilcox <willy@infradead.org>
Subject: mm-add-__dump_folio-fix-3
Date: Mon, 4 Mar 2024 19:02:43 +0000

Further testing revealed some more problems.  We were getting confused
between various different pointers leading to spurious messages about
the page not matching the folio and passing the wrong pointer to
__dump_folio().  Here's a fix-3 patch which I tested like so:

+static int __init page_dump(void)
+{
+       struct page *page;
+
+       printk("testing page dump\n");
+
+       page = alloc_page(GFP_KERNEL);
+       dump_page(page, "single");
+       put_page(page);
+       page = alloc_pages(GFP_KERNEL | __GFP_COMP, 2);
+       dump_page(page, "head");
+       dump_page(page + 1, "tail 1");
+       dump_page(page + 2, "tail 2");
+       dump_page(page + 3, "tail 3");
+       put_page(page);
+
+       return 0;
+}
+
+module_init(page_dump);

(needed some extra debug to check the values of the pointers being
passed to __dump_folio() which we wouldn't want to include)

Link: https://lkml.kernel.org/r/ZeYa00ixxC4k1ot-@casper.infradead.org
Signed-off-by: Matthew Wilcox (Oracle) <willy@infradead.org>
Cc: Aishwarya TCV <aishwarya.tcv@arm.com>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Mark Brown <broonie@kernel.org>
Cc: Ryan Roberts <ryan.roberts@arm.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 mm/debug.c |   36 ++++++++++++++++++++++--------------
 1 file changed, 22 insertions(+), 14 deletions(-)

--- a/mm/debug.c~mm-add-__dump_folio-fix-3
+++ a/mm/debug.c
@@ -121,23 +121,31 @@ static void __dump_page(const struct pag
 again:
 	memcpy(&precise, page, sizeof(*page));
 	foliop = page_folio(&precise);
-	idx = folio_page_idx(foliop, page);
-	if (idx != 0) {
-		if (idx < MAX_FOLIO_NR_PAGES) {
-			memcpy(&folio, foliop, 2 * sizeof(struct page));
-			nr_pages = folio_nr_pages(&folio);
-		}
+	if (foliop == (struct folio *)&precise) {
+		idx = 0;
+		if (!folio_test_large(foliop))
+			goto dump;
+		foliop = (struct folio *)page;
+	} else {
+		idx = folio_page_idx(foliop, page);
+	}
+
+	if (idx < MAX_FOLIO_NR_PAGES) {
+		memcpy(&folio, foliop, 2 * sizeof(struct page));
+		nr_pages = folio_nr_pages(&folio);
+		foliop = &folio;
+	}
 
-		if (idx > nr_pages) {
-			if (loops-- > 0)
-				goto again;
-			printk("page does not match folio\n");
-			precise.compound_head &= ~1UL;
-			foliop = (struct folio *)&precise;
-			idx = 0;
-		}
+	if (idx > nr_pages) {
+		if (loops-- > 0)
+			goto again;
+		printk("page does not match folio\n");
+		precise.compound_head &= ~1UL;
+		foliop = (struct folio *)&precise;
+		idx = 0;
 	}
 
+dump:
 	__dump_folio(foliop, &precise, pfn, idx);
 }
 
_

Patches currently in -mm which might be from willy@infradead.org are

mm-separate-out-folio_flags-from-pageflags.patch
mm-remove-pagewaiters-pagesetwaiters-and-pageclearwaiters.patch
mm-remove-pageyoung-and-pageidle-definitions.patch
mm-add-__dump_folio.patch
mm-make-dump_page-take-a-const-argument.patch
mm-constify-testing-page-folio-flags.patch
mm-constify-more-page-folio-tests.patch
mm-remove-cast-from-page_to_nid.patch
mm-mempolicy-use-a-folio-in-do_mbind.patch


^ permalink raw reply	[flat|nested] only message in thread

only message in thread, other threads:[~2024-03-05 20:47 UTC | newest]

Thread overview: (only message) (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-03-05 20:47 [folded-merged] mm-add-__dump_folio-fix-3.patch removed from -mm tree Andrew Morton

This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.