RE: [Intel-wired-lan] [PATCH] e1000e: prevent division by zero if TIMINCA is zero

From: Brown, Aaron F
Date: Mon May 16 2016 - 18:32:02 EST


> From: Intel-wired-lan [mailto:intel-wired-lan-bounces@xxxxxxxxxxxxxxxx] On
> Behalf Of Denys Vlasenko
> Sent: Friday, May 6, 2016 12:42 PM
> To: Kirsher, Jeffrey T <jeffrey.t.kirsher@xxxxxxxxx>
> Cc: intel-wired-lan@xxxxxxxxxxxxxxxx; Denys Vlasenko
> <dvlasenk@xxxxxxxxxx>; LKML <linux-kernel@xxxxxxxxxxxxxxx>;
> netdev@xxxxxxxxxxxxxxx
> Subject: [Intel-wired-lan] [PATCH] e1000e: prevent division by zero if
> TIMINCA is zero
>
> Users report that under VMWare, er32(TIMINCA) returns zero.
> This causes division by zero at init time as follows:
>
> ==> incvalue = er32(TIMINCA) & E1000_TIMINCA_INCVALUE_MASK;
> for (i = 0; i < E1000_MAX_82574_SYSTIM_REREADS; i++) {
> /* latch SYSTIMH on read of SYSTIML */
> systim_next = (cycle_t)er32(SYSTIML);
> systim_next |= (cycle_t)er32(SYSTIMH) << 32;
>
> time_delta = systim_next - systim;
> temp = time_delta;
> ====> rem = do_div(temp, incvalue);
>
> This change makes kernel survive this, and users report that
> NIC does work after this change.
>
> Since on real hardware incvalue is never zero, this should not affect
> real hardware use case.
>
> Signed-off-by: Denys Vlasenko <dvlasenk@xxxxxxxxxx>
> CC: Jeff Kirsher <jeffrey.t.kirsher@xxxxxxxxx>
> CC: "Ruinskiy, Dima" <dima.ruinskiy@xxxxxxxxx>
> CC: intel-wired-lan@xxxxxxxxxxxxxxxx
> CC: netdev@xxxxxxxxxxxxxxx
> CC: LKML <linux-kernel@xxxxxxxxxxxxxxx>
> ---
> drivers/net/ethernet/intel/e1000e/netdev.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)

As Mark Rustad pointed out I recall this was earlier rejected as something that is a VMWare error and it should be fixed there so that existing VMs will start working without installing a new driver. Having said that, it does not seem to be causing any harm in my testing, so...

Tested-by: Aaron Brown <aaron.f.brown@xxxxxxxxx>