From 5d4156ac48a4f4c0b7f980a51937e33362f78911 Mon Sep 17 00:00:00 2001 From: Linus Torvalds Date: Sun, 22 Sep 2019 11:31:15 -0700 Subject: firmware: bcm47xx_nvram: _really_ correct size_t printf format MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Commit feb4eb060c3a ("firmware: bcm47xx_nvram: Correct size_t printf format") was wrong, and changed a printout of 'header.len' - which is an u32 type - to use '%zu'. It apparently did pattern matching on the other case, where it printed out 'nvram_len', which is indeed of type 'size_t'. Rather than undoing the change, this just makes it use the variable that the change seemed to expect to be used. Cc: Florian Fainelli Cc: Philippe Mathieu-Daudé Cc: Paul Burton Signed-off-by: Linus Torvalds --- drivers/firmware/broadcom/bcm47xx_nvram.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'drivers/firmware') diff --git a/drivers/firmware/broadcom/bcm47xx_nvram.c b/drivers/firmware/broadcom/bcm47xx_nvram.c index 6d2820f6aca1..da04fdae62a1 100644 --- a/drivers/firmware/broadcom/bcm47xx_nvram.c +++ b/drivers/firmware/broadcom/bcm47xx_nvram.c @@ -149,7 +149,7 @@ static int nvram_init(void) nvram_len = header.len; if (nvram_len >= NVRAM_SPACE) { pr_err("nvram on flash (%zu bytes) is bigger than the reserved space in memory, will just copy the first %i bytes\n", - header.len, NVRAM_SPACE); + nvram_len, NVRAM_SPACE); nvram_len = NVRAM_SPACE - 1; } -- cgit v1.2.3-59-g8ed1b