> On Mar 27, 2020, at 11:53 AM, J. Bruce Fields <bfields@xxxxxxxxxx> wrote:
>
> From: "J. Bruce Fields" <bfields@xxxxxxxxxx>
>
> Trond points out in 277f27e2f277 that we allow invalid cache entries to
> persist indefinitely. That fix, however, reintroduces the problem fixed
> by Kinglong Mee's d6fc8821c2d2 "SUNRPC/Cache: Always treat the invalid
> cache as unexpired", where an invalid cache entry is immediately removed
> by a flush before mountd responds to it. The result is that the server
> thread that should be waiting for mountd to fill in that entry instead
> gets an -ETIMEDOUT return from cache_check(). Symptoms are the server
> becoming unresponsive after a restart, reproduceable by running pynfs
> 4.1 test REBT5.
>
> Instead, take a compromise approach: allow invalid cache entries to be
> removed after they expire, but not to be removed by a cache flush.
>
> Fixes: 277f27e2f277 "SUNRPC/cache: Allow garbage collection..."
> Signed-off-by: J. Bruce Fields <bfields@xxxxxxxxxx>
Thanks, Bruce. checkpatch.pl is complaining loudly about the style of
the short commit descriptions. I'll fix those up before applying it to
nfsd-5.7.
> ---
> include/linux/sunrpc/cache.h | 7 +++++--
> 1 file changed, 5 insertions(+), 2 deletions(-)
>
> diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h
> index 532cdbda43da..10891b70fc7b 100644
> --- a/include/linux/sunrpc/cache.h
> +++ b/include/linux/sunrpc/cache.h
> @@ -209,8 +209,11 @@ static inline void cache_put(struct cache_head *h, struct cache_detail *cd)
>
> static inline bool cache_is_expired(struct cache_detail *detail, struct cache_head *h)
> {
> - return (h->expiry_time < seconds_since_boot()) ||
> - (detail->flush_time >= h->last_refresh);
> + if (h->expiry_time < seconds_since_boot())
> + return true;
> + if (!test_bit(CACHE_VALID, &h->flags))
> + return false;
> + return detail->flush_time >= h->last_refresh;
> }
>
> extern int cache_check(struct cache_detail *detail,
> --
> 2.25.1
>
--
Chuck Lever
