[PATCH] erofs: suppress false positive last_block warning

[Gao Xiang]

As Andrew mentioned, some rare specific gcc versions could report
last_block uninitialized warning. Actually last_block doesn't need
to be uninitialized first from its implementation due to bio == NULL
condition. After a bio is allocated, last_block will be assigned
then.

The detailed analysis is in this thread [1]. So let's silence those
confusing gccs simply.

[1] https://lore.kernel.org/r/20200421072839.GA13867@hsiangkao-HP-ZHAN-66-Pro-G1

Cc: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
Signed-off-by: Gao Xiang <hsiangkao@xxxxxxxxxx>
---
 fs/erofs/data.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/fs/erofs/data.c b/fs/erofs/data.c
index fc3a8d8064f8..2812645b361e 100644
--- a/fs/erofs/data.c
+++ b/fs/erofs/data.c
@@ -265,7 +265,7 @@ static inline struct bio *erofs_read_raw_page(struct bio *bio,
  */
 static int erofs_raw_access_readpage(struct file *file, struct page *page)
 {
-	erofs_off_t last_block;
+	erofs_off_t uninitialized_var(last_block);
 	struct bio *bio;
 
 	trace_erofs_readpage(page, true);
@@ -285,7 +285,7 @@ static int erofs_raw_access_readpages(struct file *filp,
 				      struct list_head *pages,
 				      unsigned int nr_pages)
 {
-	erofs_off_t last_block;
+	erofs_off_t uninitialized_var(last_block);
 	struct bio *bio = NULL;
 	gfp_t gfp = readahead_gfp_mask(mapping);
 	struct page *page = list_last_entry(pages, struct page, lru);
-- 
2.18.1