[coreboot-gerrit] New patch to review for coreboot: b2bfce7 x86: remove printk() from init_timer()

Aaron Durbin (adurbin@chromium.org) gerrit at coreboot.org
Tue Jun 9 22:09:06 CEST 2015

Aaron Durbin (adurbin at chromium.org) just uploaded a new patch set to gerrit, which you can find at http://review.coreboot.org/10482


commit b2bfce716becd49391177a6af13aa91c25454ee4
Author: Aaron Durbin <adurbin at chromium.org>
Date:   Tue Jun 9 15:02:50 2015 -0500

    x86: remove printk() from init_timer()
    For console drivers which use udelay() we can deadlock
    in the printk path on the spinlock. The reason is that
    on the first call to udelay() from within a console driver
    it will go back down the printk() path deadlocking oneself.
    Just remove the printk() as it was asymmetric on romstage
    vs ramstage.
    Change-Id: Iba212e3a7f249087eef0fa1cd7ad4817cb5947d5
    Signed-off-by: Aaron Durbin <adurbin at chromium.org>
 src/cpu/x86/tsc/delay_tsc.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/src/cpu/x86/tsc/delay_tsc.c b/src/cpu/x86/tsc/delay_tsc.c
index 5aaba90..0ad5d3b 100644
--- a/src/cpu/x86/tsc/delay_tsc.c
+++ b/src/cpu/x86/tsc/delay_tsc.c
@@ -152,10 +152,8 @@ static unsigned long long calibrate_tsc(void)
 void init_timer(void)
-	if (!clocks_per_usec) {
+	if (!clocks_per_usec)
 		clocks_per_usec = calibrate_tsc();
-		printk(BIOS_INFO, "clocks_per_usec: %lu\n", clocks_per_usec);
-	}
 static inline unsigned long get_clocks_per_usec(void)

More information about the coreboot-gerrit mailing list