ARM: udelay: prevent math rounding resulting in short udelays



We perform the microseconds to loops calculation using a number of
multiplies and shift rights.  Each shift right rounds down the
resulting value, which can result in delays shorter than requested.
Ensure that we always round up.

Signed-off-by: Russell King <rmk+kernel@arm.linux.org.uk>

parent 9e9bc973

arch/arm/lib/delay.S

+4 −0

Original line number	Diff line number	Diff line
		@@ -25,11 +25,15 @@ ENTRY(__udelay)
		ldr r2, .LC1
		mul r0, r2, r0
		ENTRY(__const_udelay) @ 0 <= r0 <= 0x7fffff06
		mov r1, #-1
		ldr r2, .LC0
		ldr r2, [r2] @ max = 0x01ffffff
		add r0, r0, r1, lsr #32-14
		mov r0, r0, lsr #14 @ max = 0x0001ffff
		add r2, r2, r1, lsr #32-10
		mov r2, r2, lsr #10 @ max = 0x00007fff
		mul r0, r2, r0 @ max = 2^32-1
		add r0, r0, r1, lsr #32-6
		movs r0, r0, lsr #6
		moveq pc, lr