Probe goroutine create and exit for request time

[grcevski]

With this PR we add two new uprobes, one at runtime.newproc1 and another at runtime.goexit1. When we capture the time at runtime.newproc1, we can tell the total request time, not just the service time.

I added nested spans in the traces to capture this information. When testing with our 10,000 QPS load driver, I was able to see pathological cases where the request time was dominated by the time the request waits in queue.

This PR is based on #46. It needs to be merged first.

[codecov-commenter]

Codecov Report

Merging #49 (df56dbc) into main (d41cb3e) will increase coverage by 0.10%.
The diff coverage is 44.57%.

@@            Coverage Diff             @@
##             main      #49      +/-   ##
==========================================
+ Coverage   51.13%   51.23%   +0.10%     
==========================================
  Files          14       14              
  Lines         927      968      +41     
==========================================
+ Hits          474      496      +22     
- Misses        418      435      +17     
- Partials       35       37       +2     

Flag	Coverage Δ
unittests	51.23% <44.57%> (+0.10%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
pkg/ebpf/nethttp/nethttp.go	0.00% <0.00%> (ø)
pkg/export/debug/debug.go	15.38% <0.00%> (-0.62%)	⬇️
pkg/pipe/config.go	53.33% <ø> (ø)
pkg/export/otel/metrics.go	80.68% <100.00%> (ø)
pkg/export/otel/traces.go	80.76% <100.00%> (+4.90%)	⬆️
pkg/pipe/instrumenter.go	83.33% <100.00%> (+0.72%)	⬆️
pkg/transform/spanner.go	81.33% <100.00%> (+0.51%)	⬆️

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

[grcevski]

Thanks Mario!