Re: FS-Cache: Add a helper to bulk uncache pages on an inode

Re: FS-Cache: Add a helper to bulk uncache pages on an inode

[Geert Uytterhoeven]

On Thu, Jul 7, 2011 at 22:59, Linux Kernel Mailing List
<linux-kernel@vger.kernel.org> wrote:
>    FS-Cache: Add a helper to bulk uncache pages on an inode
>
>    Add an FS-Cache helper to bulk uncache pages on an inode.  This will
>    only work for the circumstance where the pages in the cache correspond
>    1:1 with the pages attached to an inode's page cache.
>
>    This is required for CIFS and NFS: When disabling inode cookie, we were
>    returning the cookie and setting cifsi->fscache to NULL but failed to
>    invalidate any previously mapped pages.  This resulted in "Bad page
>    state" errors and manifested in other kind of errors when running
>    fsstress.  Fix it by uncaching mapped pages when we disable the inode
>    cookie.
>
>    This patch should fix the following oops and "Bad page state" errors
>    seen during fsstress testing.

> --- a/fs/fscache/page.c
> +++ b/fs/fscache/page.c
> @@ -954,3 +954,47 @@ void fscache_mark_pages_cached(struct fscache_retrieval *op,
>        pagevec_reinit(pagevec);
>  }
>  EXPORT_SYMBOL(fscache_mark_pages_cached);
> +
> +/*
> + * Uncache all the pages in an inode that are marked PG_fscache, assuming them
> + * to be associated with the given cookie.
> + */
> +void __fscache_uncache_all_inode_pages(struct fscache_cookie *cookie,
> +                                      struct inode *inode)
> +{
> +       struct address_space *mapping = inode->i_mapping;
> +       struct pagevec pvec;
> +       pgoff_t next;
> +       int i;
> +
> +       _enter("%p,%p", cookie, inode);
> +
> +       if (!mapping || mapping->nrpages == 0) {
> +               _leave(" [no pages]");
> +               return;
> +       }
> +
> +       pagevec_init(&pvec, 0);
> +       next = 0;
> +       while (next <= (loff_t)-1 &&

On m68k, I get:

fs/fscache/page.c: In function ‘__fscache_uncache_all_inode_pages’:
fs/fscache/page.c:979: warning: comparison is always false due to
limited range of data type

next is pgoff_t, which is defined to unsigned long by default, unless
overridden by the arch
(hmm, no single arch seems to do that?).
loff_t is (signed) long long.

> +              pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE)
> +              ) {
> +               for (i = 0; i < pagevec_count(&pvec); i++) {
> +                       struct page *page = pvec.pages[i];
> +                       pgoff_t page_index = page->index;
> +
> +                       ASSERTCMP(page_index, >=, next);
> +                       next = page_index + 1;
> +
> +                       if (PageFsCache(page)) {
> +                               __fscache_wait_on_page_write(cookie, page);
> +                               __fscache_uncache_page(cookie, page);
> +                       }
> +               }
> +               pagevec_release(&pvec);
> +               cond_resched();
> +       }
> +
> +       _leave("");
> +}

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

[PATCH] FS-Cache: Fix bounds check

[David Howells]

__fscache_uncache_all_inode_pages() has a loop that goes through page index
numbers go up to (loff_t)-1.  This is incorrect.  The limit should be
(pgoff_t)-1 as on a 32-bit machine the pgoff_t is smaller than loff_t.

On m68k the following error is observed:

fs/fscache/page.c: In function '__fscache_uncache_all_inode_pages':
fs/fscache/page.c:979: warning: comparison is always false due to
limited range of data type

[Should there be a PGOFF_T_MAX constant defined?]

Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: David Howells <dhowells@redhat.com>
---

 fs/fscache/page.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/fs/fscache/page.c b/fs/fscache/page.c
index 60315b3..112359d 100644
--- a/fs/fscache/page.c
+++ b/fs/fscache/page.c
@@ -993,7 +993,7 @@ void __fscache_uncache_all_inode_pages(struct fscache_cookie *cookie,
 
 	pagevec_init(&pvec, 0);
 	next = 0;
-	while (next <= (loff_t)-1 &&
+	while (next <= (pgoff_t)-1 &&
 	       pagevec_lookup(&pvec, mapping, next, PAGEVEC_SIZE)
 	       ) {
 		for (i = 0; i < pagevec_count(&pvec); i++) {

Re: [stable] [PATCH] FS-Cache: Fix bounds check

[Greg KH]

On Wed, Jul 13, 2011 at 12:55:18PM +0100, David Howells wrote:
> __fscache_uncache_all_inode_pages() has a loop that goes through page index
> numbers go up to (loff_t)-1.  This is incorrect.  The limit should be
> (pgoff_t)-1 as on a 32-bit machine the pgoff_t is smaller than loff_t.
> 
> On m68k the following error is observed:
> 
> fs/fscache/page.c: In function '__fscache_uncache_all_inode_pages':
> fs/fscache/page.c:979: warning: comparison is always false due to
> limited range of data type
> 
> [Should there be a PGOFF_T_MAX constant defined?]
> 
> Reported-by: Geert Uytterhoeven <geert@linux-m68k.org>
> Signed-off-by: David Howells <dhowells@redhat.com>
> ---
> 
>  fs/fscache/page.c |    2 +-
>  1 files changed, 1 insertions(+), 1 deletions(-)


<formletter>

This is not the correct way to submit patches for inclusion in the
stable kernel tree.  Please read Documentation/stable_kernel_rules.txt
for how to do this properly.

</formletter>