> -----Original Message-----
> From: David Ahern <dsah...@gmail.com>
> Sent: Tuesday, October 06, 2020 10:56 PM
> To: Keller, Jacob E <jacob.e.kel...@intel.com>; netdev@vger.kernel.org; Jakub
> Kicinski <k...@kernel.org>; Shannon Nelson <snel...@pensando.io>
> Subject: Re: [iproute2-next v2 1/1] devlink: display elapsed time during flash
> update
>
> On 9/30/20 4:40 PM, Jacob Keller wrote:
> > @@ -3124,12 +3140,19 @@ static int cmd_dev_flash_status_cb(const struct
> nlmsghdr *nlh, void *data)
> > done =
> mnl_attr_get_u64(tb[DEVLINK_ATTR_FLASH_UPDATE_STATUS_DONE]);
> > if (tb[DEVLINK_ATTR_FLASH_UPDATE_STATUS_TOTAL])
> > total =
> mnl_attr_get_u64(tb[DEVLINK_ATTR_FLASH_UPDATE_STATUS_TOTAL]);
> > + if (tb[DEVLINK_ATTR_FLASH_UPDATE_STATUS_TIMEOUT])
> > + ctx->status_msg_timeout =
> mnl_attr_get_u64(tb[DEVLINK_ATTR_FLASH_UPDATE_STATUS_TIMEOUT]);
> > + else
> > + ctx->status_msg_timeout = 0;
> >
> > if (!nullstrcmp(msg, ctx->last_msg) &&
> > !nullstrcmp(component, ctx->last_component) &&
> > ctx->last_pc && ctx->not_first) {
> > pr_out_tty("\b\b\b\b\b"); /* clean percentage */
> > } else {
> > + /* only update the last status timestamp if the message changed
> */
> > + gettimeofday(&ctx->time_of_last_status, NULL);
>
> gettimeofday/REALCLOCK should not be used for measuring time differences.
>
Ah. Good point. This should use CLOCK_MONOTONIC instead, right?
> > +
> > if (ctx->not_first)
> > pr_out("\n");
> > if (component) {
> > @@ -3155,11 +3178,72 @@ static int cmd_dev_flash_status_cb(const struct
> nlmsghdr *nlh, void *data)
> > return MNL_CB_STOP;
> > }
> >
> > +static void cmd_dev_flash_time_elapsed(struct cmd_dev_flash_status_ctx
> *ctx)
> > +{
> > + struct timeval now, res;
> > +
> > + gettimeofday(&now, NULL);
> > + timersub(&now, &ctx->time_of_last_status, &res);
> > +
