6 Commits (jonathan/netmon-changedelta)

Author	SHA1	Message	Date
Jonathan Nobels	db7c05aa59	ipn/ipnlocal, net/netmon: remove ChangeDelta.New and Old ... updates tailscale/corp#33891 Signed-off-by: Jonathan Nobels <jonathan@tailscale.com>	1 day ago
Jonathan Nobels	696191f850	net/netmon: remove ChangeDelta.Major and start the TODOs ... updates tailscale/corp#33891 We've had some TODO's in netmon for ages related to asking more pointed questions of ChangeDelta. We're also seeing a lot of ChangeDelta's being flagged as "Major" when they are not interesting, triggering rebinds in wgengine that are not needed. This removes the "Major" flag and replaces it with a set of precomputed values that hint at what the ChangeDelta means. The dependencies are cleaned up a bit, notably removing dependency on netmon itself for calculating what is interesting, and what is not. This includes letting individual platforms set a bespoke global "IsInterestingInterface" function. This was only used on Darwin. RebindRequired roughly flows from how "Major" was historically calculated but includes some additional checks for various uninteresting events such as changes in interface addresses that shouldn't trigger a rebind to minimize thrashing. The individual values that we roll into RebindRequired are also exposed so that components consuming netmap.ChangeDelta can ask more targeted questions. Signed-off-by: Jonathan Nobels <jonathan@tailscale.com> wgengine/userspace: skip rebind and restun if all of our interfaces are down updates tailscale/corp#33891 Skip rebind and restun if we have called SetNetworkUp with false. Neither cares about this state and it's just flappy to try to reStun and rebind if there are no usable interfaces available. Signed-off-by: Jonathan Nobels <jonathan@tailscale.com> Add viability flag and fix bug in state	1 day ago
M. J. Fromberger	ab4b990d51	net/netmon: do not abandon a subscriber when exiting early (#17899) ... LinkChangeLogLimiter keeps a subscription to track rate limits for log messages. But when its context ended, it would exit the subscription loop, leaving the subscriber still alive. Ensure the subscriber gets cleaned up when the context ends, so we don't stall event processing. Updates tailscale/corp#34311 Change-Id: I82749e482e9a00dfc47f04afbc69dd0237537cb2 Signed-off-by: M. J. Fromberger <fromberger@tailscale.com>	3 weeks ago
Claus Lensbøl	ce752b8a88	net/netmon: remove usage of direct callbacks from netmon (#17292) ... The callback itself is not removed as it is used in other repos, making it simpler for those to slowly transition to the eventbus. Updates #15160 Signed-off-by: Claus Lensbøl <claus@tailscale.com>	2 months ago
Brad Fitzpatrick	8b48f3847d	net/netmon, wgengine/magicsock: simplify LinkChangeLogLimiter signature ... Remove the need for the caller to hold on to and call an unregister function. Both two callers (one real, one test) already have a context they can use. Use context.AfterFunc instead. There are no observable side effects from scheduling too late if the goroutine doesn't run sync. Updates #17148 Change-Id: Ie697dae0e797494fa8ef27fbafa193bfe5ceb307 Signed-off-by: Brad Fitzpatrick <bradfitz@tailscale.com>	3 months ago
Andrew Dunham	640b2fa3ae	net/netmon, wgengine/magicsock: be quieter with portmapper logs ... This adds a new helper to the netmon package that allows us to rate-limit log messages, so that they only print once per (major) LinkChange event. We then use this when constructing the portmapper, so that we don't keep spamming logs forever on the same network. Updates #13145 Signed-off-by: Andrew Dunham <andrew@du.nham.ca> Change-Id: I6e7162509148abea674f96efd76be9dffb373ae4	9 months ago