[PATCH -mm] mm: readahead: apply a default readahead size

[Randy Dunlap]

UBSAN reports an invalid shift size:

mr-fox kernel: UBSAN: shift-out-of-bounds in ./include/linux/log2.h:57:13
mr-fox kernel: shift exponent 64 is too large for 64-bit type 'long unsigned int'

Original report:
https://lore.kernel.org/lkml/c6e5eb81-680f-dd5c-8a81-62041a5ce50c@xxxxxx/

Follow-up report:
https://lore.kernel.org/lkml/0c283ea9-b446-0e40-6dc8-e9585ae171b4@xxxxxx/T/#m9b604660925f9e8a544f7453130c31d083c1e5bb


Willy suggested that get_init_ra_size() was being called with a size of 0,
which would cause this (instead of some Huge value), so add a check in
that function for size == 0, and if 0, default it to 32 (pages).

Reported-by: Toralf Förster <toralf.foerster@xxxxxx>
Signed-off-by: Randy Dunlap <rdunlap@xxxxxxxxxxxxx>
Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Cc: Jens Axboe <axboe@xxxxxxxxx>
Cc: Matthew Wilcox <willy@xxxxxxxxxxxxx>
Cc: linux-mm@xxxxxxxxx
---
 mm/readahead.c |    6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)

--- linux-5.10.1.orig/mm/readahead.c
+++ linux-5.10.1/mm/readahead.c
@@ -310,7 +310,11 @@ void force_page_cache_ra(struct readahea
  */
 static unsigned long get_init_ra_size(unsigned long size, unsigned long max)
 {
-	unsigned long newsize = roundup_pow_of_two(size);
+	unsigned long newsize;
+
+	if (!size)
+		size = 32;
+	newsize = roundup_pow_of_two(size);
 
 	if (newsize <= max / 32)
 		newsize = newsize * 4;