Hi Stefan, Jakub,
kuba@xxxxxxxxxx wrote on Fri, 3 Feb 2023 20:19:23 -0800:
> On Tue, 29 Nov 2022 17:00:41 +0100 Miquel Raynal wrote:
> > +static int nl802154_trigger_scan(struct sk_buff *skb, struct genl_info *info)
> > +{
> > + struct cfg802154_registered_device *rdev = info->user_ptr[0];
> > + struct net_device *dev = info->user_ptr[1];
> > + struct wpan_dev *wpan_dev = dev->ieee802154_ptr;
> > + struct wpan_phy *wpan_phy = &rdev->wpan_phy;
> > + struct cfg802154_scan_request *request;
> > + u8 type;
> > + int err;
> > +
> > + /* Monitors are not allowed to perform scans */
> > + if (wpan_dev->iftype == NL802154_IFTYPE_MONITOR)
>
> extack ?
Thanks for pointing at it, I just did know about it. I did convert
most of the printk's into extack strings. Shall I keep both or is fine
to just keep the extack thing?
For now I've dropped the printk's, please tell me if this is wrong.
>
> > + return -EPERM;
Stefan, do you prefer a series of patches applying on top of your
current next or should I re-roll the entire series (scan + beacons)?
I am preparing a series applying on top of the current list of applied
patches. This means next PR to net maintainers will include this patch
as it is today + fixes on top. If this is fine for both parties, I will
send these (including the other changes discussed with Alexander). Just
let me know.
Sorry btw for the delay, I really had to finish other activities before
switching back.
> > +
> > + request = kzalloc(sizeof(*request), GFP_KERNEL);
> > + if (!request)
> > + return -ENOMEM;
> > +
> > + request->wpan_dev = wpan_dev;
> > + request->wpan_phy = wpan_phy;
> > +
> > + type = nla_get_u8(info->attrs[NL802154_ATTR_SCAN_TYPE]);
>
> what checks info->attrs[NL802154_ATTR_SCAN_TYPE] is not NULL?
>
> > + switch (type) {
> > + case NL802154_SCAN_PASSIVE:
> > + request->type = type;
> > + break;
> > + default:
> > + pr_err("Unsupported scan type: %d\n", type);
> > + err = -EINVAL;
>
> extack (printfs are now supported)
>
> > + goto free_request;
> > + }
> > +
> > + if (info->attrs[NL802154_ATTR_PAGE]) {
> > + request->page = nla_get_u8(info->attrs[NL802154_ATTR_PAGE]);
> > + if (request->page > IEEE802154_MAX_PAGE) {
>
> bound check should be part of the policy NLA_POLICY_MAX()
I just improved the policies to make these checks useless and simplify a
lot the code there, thanks as well for pointing at it.
> > + pr_err("Invalid page %d > %d\n",
> > + request->page, IEEE802154_MAX_PAGE);
> > + err = -EINVAL;
>
> extack
>
> > + goto free_request;
> > + }
> > + } else {
> > + /* Use current page by default */
> > + request->page = wpan_phy->current_page;
> > + }
> > +
> > + if (info->attrs[NL802154_ATTR_SCAN_CHANNELS]) {
> > + request->channels = nla_get_u32(info->attrs[NL802154_ATTR_SCAN_CHANNELS]);
> > + if (request->channels >= BIT(IEEE802154_MAX_CHANNEL + 1)) {
>
> policy as well
>
> > + pr_err("Invalid channels bitfield %x ≥ %lx\n",
> > + request->channels,
> > + BIT(IEEE802154_MAX_CHANNEL + 1));
> > + err = -EINVAL;
> > + goto free_request;
> > + }
> > + } else {
> > + /* Scan all supported channels by default */
> > + request->channels = wpan_phy->supported.channels[request->page];
> > + }
> > +
> > + if (info->attrs[NL802154_ATTR_SCAN_PREAMBLE_CODES] ||
> > + info->attrs[NL802154_ATTR_SCAN_MEAN_PRF]) {
> > + pr_err("Preamble codes and mean PRF not supported yet\n");
>
> NLA_REJECT also in policy
>
> > + err = -EINVAL;
> > + goto free_request;
> > + }
> > +
> > + if (info->attrs[NL802154_ATTR_SCAN_DURATION]) {
> > + request->duration = nla_get_u8(info->attrs[NL802154_ATTR_SCAN_DURATION]);
> > + if (request->duration > IEEE802154_MAX_SCAN_DURATION) {
> > + pr_err("Duration is out of range\n");
> > + err = -EINVAL;
> > + goto free_request;
> > + }
> > + } else {
> > + /* Use maximum duration order by default */
> > + request->duration = IEEE802154_MAX_SCAN_DURATION;
> > + }
> > +
> > + if (wpan_dev->netdev)
> > + dev_hold(wpan_dev->netdev);
>
> Can we put a tracker in the request and use netdev_hold() ?
I'll look into it.
Thanks,
Miquèl
