Codechange: encapsulate network error handling (and fix #9142) #9146
+238
−136
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
rubidium42
changed the title
rubidium42
added
the
backport requested
TrueBrain
reviewed
Apr 30, 2021
There was a problem hiding this comment.
I think this is fine; I only wonder if we shouldn't just put this in the header-file, so we can let compilers optimize it away immediately. Otherwise I guess it needs some LTO to do it properly. No clue if it matters either way.
Also, I would like a second opinion on the C++ side of this, but yeah .. I think this is the better solution to go with, as we don't force every OS into a Linux-like experience :D
rubidium42
force-pushed
the
network_error_handling
branch
from
d49abc7 to
e30e5a7
Compare
TrueBrain
reviewed
Apr 30, 2021
LordAro
reviewed
Apr 30, 2021
rubidium42
force-pushed
the
network_error_handling
branch
from
e30e5a7 to
c2989c2
Compare
ccapitalK
reviewed
May 1, 2021
|
I realise my previous comments somewhat contradict this, but given (almost?) every use of NetworkError::AsString/GetLastAsString is in DEBUG() that need a const char*... maybe that would be a better return type? :> Would save the very noisy c_str() call every time... |
rubidium42
force-pushed
the
network_error_handling
branch
from
8ccbd97 to
145d128
Compare
LordAro
added
backported
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Closes #9142.
Motivation / Problem
The discussion about #9142 and how to solve that nicely (and permanently) yielded another rabbit hole.
This does solve #9142, but it should also prevent it from happening again as doing it another way would break the compile for at least Windows.
Description
Create a NetworkError class that has functions to query what kind of error it is. This way the actual definition of what a particular error entails needs to be coded only once. #9142 is a nice example of why the current method is not sustainable; if someone else adds a new place where recv/send is used, and would just use
err == EWOULDBLOCKthen #9142 happens again. By abstracting the whole error type determination to a single location, it needs to be set only once and can then be reused.To not have huge amounts of complicated #ifdefs in os_abstraction.h, the actual implementation has been placed in os_abstraction.cpp and some of the functions defined in os_abstraction.h have been moved as well.
By this refactor the #ifdefs for the functionality are also moved into the functions themselves, so for one function you can easily see the implementation for the different environments, whereas you had to really look well through os_abstraction.h before to find out what would happen exactly in different environments.
Limitations
None, as far as I know.
Checklist for review
Some things are not automated, and forgotten often. This list is a reminder for the reviewers.