[PATCH AUTOSEL 5.10 25/33] ptp_qoriq: fix overflow in ptp_qoriq_adjfine() u64 calcalation
From: Sasha Levin <sashal@kernel.org>
Date: 2021-03-29 22:27:17
Also in:
lkml, stable
Subsystem:
freescale qoriq ptp clock driver, networking drivers, ptp hardware clock support, the rest · Maintainers:
Yangbo Lu, Andrew Lunn, "David S. Miller", Eric Dumazet, Jakub Kicinski, Paolo Abeni, Richard Cochran, Linus Torvalds
From: Yangbo Lu <yangbo.lu@nxp.com> [ Upstream commit f51d7bf1dbe5522c51c93fe8faa5f4abbdf339cd ] Current calculation for diff of TMR_ADD register value may have 64-bit overflow in this code line, when long type scaled_ppm is large. adj *= scaled_ppm; This patch is to resolve it by using mul_u64_u64_div_u64(). Signed-off-by: Yangbo Lu <yangbo.lu@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net> Signed-off-by: Sasha Levin <sashal@kernel.org> --- drivers/ptp/ptp_qoriq.c | 13 +++++++------ 1 file changed, 7 insertions(+), 6 deletions(-)
diff --git a/drivers/ptp/ptp_qoriq.c b/drivers/ptp/ptp_qoriq.c
index beb5f74944cd..08f4cf0ad9e3 100644
--- a/drivers/ptp/ptp_qoriq.c
+++ b/drivers/ptp/ptp_qoriq.c@@ -189,15 +189,16 @@ int ptp_qoriq_adjfine(struct ptp_clock_info *ptp, long scaled_ppm) tmr_add = ptp_qoriq->tmr_add; adj = tmr_add; - /* calculate diff as adj*(scaled_ppm/65536)/1000000 - * and round() to the nearest integer + /* + * Calculate diff and round() to the nearest integer + * + * diff = adj * (ppb / 1000000000) + * = adj * scaled_ppm / 65536000000 */ - adj *= scaled_ppm; - diff = div_u64(adj, 8000000); - diff = (diff >> 13) + ((diff >> 12) & 1); + diff = mul_u64_u64_div_u64(adj, scaled_ppm, 32768000000); + diff = DIV64_U64_ROUND_UP(diff, 2); tmr_add = neg_adj ? tmr_add - diff : tmr_add + diff; - ptp_qoriq->write(®s->ctrl_regs->tmr_add, tmr_add); return 0;
--
2.30.1