What type of PR is this?

/kind bug

What this PR does / why we need it:

This PR fixes race condition:

• apiserver request times out
• We try to log timed out request and get UserAgent() from http.request - this is actually implemented as getting element from Headers map - https://cs.opensource.google/go/go/+/refs/tags/go1.17.2:src/net/http/request.go;l=406;drc=e6dda19888180c5159460486d30c0412e4980748
• Underlying goroutine can still be executing request:

  kubernetes/staging/src/k8s.io/apiserver/pkg/server/filters/timeout.go

  Line 96 in a78e313

  go func() {

• This goroutine can be modifying request's headers, for example in Authentication filter:

  kubernetes/staging/src/k8s.io/apiserver/pkg/endpoints/filters/authentication.go

  Line 77 in a78e313

  req.Header.Del("Authorization")

This can cause unrecoverable panic that crashes apiserver(this stacktrace was based on 1.21 version):

goroutine 943732 [running]:
runtime.throw
        /usr/local/go/src/runtime/panic.go:1117
runtime.mapaccess1_faststr
        /usr/local/go/src/runtime/map_faststr.go:21
net/textproto.MIMEHeader.Get
        /usr/local/go/src/net/textproto/header.go:34
net/http.Header.Get(...)
        /usr/local/go/src/net/http/header.go:46
net/http.(*Request).UserAgent(...)
k8s.io/kubernetes/vendor/k8s.io/apiserver/pkg/server/httplog.(*respLogger).LogArgs(...)
        k8s.io/kubernetes/vendor/k8s.io/apiserver/pkg/server/httplog/httplog.go:185 

I've also looked briefly and found two other possible race conditions when accessing/modifying addedInfo or addedKeyValuePairs (although I haven't observed such error in real cluster as compared to this panic above).

Which issue(s) this PR fixes:

Fixes #105759

Special notes for your reviewer:

Does this PR introduce a user-facing change?

Fix concurrent map access causing panics when logging timed-out API calls.

Additional documentation e.g., KEPs (Kubernetes Enhancement Proposals), usage docs, etc.: