From mboxrd@z Thu Jan 1 00:00:00 1970 From: linux@arm.linux.org.uk (Russell King - ARM Linux) Date: Tue, 1 Jun 2010 18:54:03 +0100 Subject: incorrect panic("failed to get clock for system timer"); ?? In-Reply-To: <4C0539E4.9090501@gmail.com> References: <4C0539E4.9090501@gmail.com> Message-ID: <20100601175403.GA6644@n2100.arm.linux.org.uk> To: linux-arm-kernel@lists.infradead.org List-Id: linux-arm-kernel.lists.infradead.org On Tue, Jun 01, 2010 at 12:48:36PM -0400, f. achkar wrote: > Hi, > Newbie debugging Linux-2.6.32-9 on S3C2450, the kernel seems to panic at : > s3c2410_timer_init () at arch/arm/plat-s3c/time.c:256 > > if (IS_ERR(timerclk)) > panic("failed to get clock for system timer"); > > more precisely 'if (IS_ERR(timerclk))' is non-null !?! although the > 'timerclk' seem to be fine: > (gdb) print *timerclk > $27 = {list = {next = 0xc04a933c, prev = 0xc04a93ac}, owner = 0x0, > parent = 0xc04abfd4, name = 0xc042c5a1 "timers", id = -1, > usage = 0, rate = 0, ctrlbit = 1024, enable = 0xc009631c > , set_rate = 0, get_rate = 0, > round_rate = 0, set_parent = 0} > (gdb) s > 257 panic("failed to get clock for system timer"); > (gdb) print (unsigned long)(timerclk) > $28 = 3226112884 > (gdb) > > I'm not sure why 'IS_ERR(timerclk)' is holding true? or is it a bug ? > thanks, Looks to me like a compiler bug. Your timerclk is 0xC04A9374, which certainly isn't a value which IS_ERR() should return true for.