Bump Go version used in v0.34.x to v1.20
#1351
Conversation
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
Signed-off-by: Thane Thomson <connect@thanethomson.com>
|
In order to get the linter to actually work on v0.34.x, I had to update the v0.34.x CI and code to use the latest golangci-lint, update the linter config, and add some ignores for specific lints in certain places. If we're going to maintain the v0.34.x branch for much longer, we should possibly consider re-enabling some of the linters I've disabled and fix the lints (a trivial, but time-consuming and low-impact job). |
p2p/conn/connection.go
Outdated
| SendQueueSize: int(atomic.LoadInt32(&channel.sendQueueSize)), | ||
| SendQueueSize: int(atomic.LoadInt32(&channel.sendQueueSize)), //nolint:gosec | ||
| Priority: channel.desc.Priority, | ||
| RecentlySent: atomic.LoadInt64(&channel.recentlySent), | ||
| RecentlySent: atomic.LoadInt64(&channel.recentlySent), //nolint:gosec |
There was a problem hiding this comment.
I had to manually skip the gosec linter checks here, otherwise it complains:
G601: Implicit memory aliasing in for loop.
My understanding of how this code works is that we're immediately loading the int values from the pointers during the current iteration, so it's irrelevant. Please correct me if I'm wrong though.
There was a problem hiding this comment.
I think it is complaining about the loop variable (potentially) being rewritten while using it as a reference. IICR @jmalicevic fixed this in other branches
Follows from the govulncheck failure in #1342, since Go v1.19 has reached EOL.
PR checklist
.changelog(we use unclog to manage our changelog)docs/orspec/) and code comments