cmd/docker: add cause to user-terminated context.Context

[Benehiko]

This patch adds a "cause" to the context.Context error when the user terminates the process through SIGINT/SIGTERM.

This allows us to distinguish the cause of the
context.Context cancellation. In future we would also be able to improve the UX of printed errors
based on the underlying cause.

- What I did

- How I did it

- How to verify it

- Description for the changelog

`docker` cli now sets the corresponding exit code when the process was killed with a signal

- A picture of a cute animal (not mandatory but encouraged)

[codecov-commenter]

Codecov Report

Attention: Patch coverage is 57.89474% with 16 lines in your changes missing coverage. Please review.

Project coverage is 59.43%. Comparing base (f97ec69) to head (c51be77).
Report is 105 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5760      +/-   ##
==========================================
- Coverage   59.47%   59.43%   -0.04%     
==========================================
  Files         346      347       +1     
  Lines       29367    29431      +64     
==========================================
+ Hits        17465    17492      +27     
- Misses      10929    10965      +36     
- Partials      973      974       +1     

[fayazkhan121]

The test looks solid. TestUserTerminatedError test case properly introduces notifyContext and handles the termination signals as expected.

[Benehiko]

Please don't merge yet, this is incorrect. We need to have status 130 print when a container exits with that code.

Spam / AI account (blocked now)

[vvoland]

LGTM overall, but we probably still need to adjust the output, see:

Before

$ docker run --rm -it --pull=always fedora
latest: Pulling from library/fedora
e3c408ecb0c2: Downloading [====================================>              ]  41.94MB/57.21MB
^Cdocker: context canceled.
See 'docker run --help'.

After

$  ./build/docker run --rm -it --pull=always fedora
latest: Pulling from library/fedora
e3c408ecb0c2: Downloading [======================>                            ]  26.21MB/57.21MB
^Cdocker: 
Run 'docker run --help' for more information

IMO, it shouldn't print anything at all, so it should look like:

$ docker run --rm -it --pull=always fedora
latest: Pulling from library/fedora
e3c408ecb0c2: Downloading [====================================>              ]  41.94MB/57.21MB
^C

[Benehiko]

@vvoland we won't print anything when this PR #5778 rebases with these changes

[vvoland]

LGTM

[laurazard]

Just a small nit/suggestion, not necessary.

You could do:

Suggested change

	type errCtxSignalTerminated struct {
	signal os.Signal
	}
	func (e errCtxSignalTerminated) Error() string {
	return ""
	}
	type errCtxSignalTerminated struct {
	signal os.Signal
	}
	func (e *errCtxSignalTerminated) Error() string {
	return ""
	}
	func (e *errCtxSignalTerminated) exitCode() int {
	if e == nil {
	return 1
	}
	s, ok := e.signal.(syscall.Signal)
	if !ok {
	return 1
	}
	return 128 + int(s)
	}

Then errCtxSignalTerminated implements error (instead of *errCtxSignalTerminated implementing error.

[laurazard]

WDYT @Benehiko @vvoland ?

[vvoland]

+1 on the exitCode

[Benehiko]

The suggestion looks alright to me, although we only use the exit code inside the getExitCode function and this error type is only used inside cmd/docker/docker.go. At least to me, it just seems like extra refactoring with little benefit.