Getting rid of Errno and WinError by waj · Pull Request #8885 · crystal-lang/crystal

waj · 2020-03-06T15:18:54Z

This PR removes the usage of Errno and WinError as exceptions and instead provides some new classes to distinguish different kind of errors.

This is the hierarchy of the new or updated exceptions:

Exception
+- RuntimeError
+- IO::Error
   +- IO::TimeoutError
   +- File::Error
   |  +- File::NotFoundError
   |  +- File::AccessDeniedError
   |  +- File::AlreadyExistsError
   +- Socket::Error
      +- Socket::ConnectError
      +- Socket::BindError

More subclasses might be added in the future but I think this is a good starting point.

Exception now has an optional os_error containing the original error code that anyone can use in case the exception class hierarchy is not enough for proper error handling
IO::Timeout is renamed to IO::TimeoutError
Errno and Winerror are now enums
Exception.from_errno and Exception.from_winerror eases creation of exceptions from OS errors
RuntimeError is used, often when there is a system call that shouldn't fail in normal conditions

(closes #8827)

straight-shoota · 2020-03-06T16:30:22Z

I don't like having os_error on Exception, spilling into the generic exception namespace when it's use is actually limited to a very small amount of exception types from stdlib and maybe a few exceptions in shards with C bindings. This propert doesn't make sense for the vast majority of exceptions in a typical Crystal program.
Can we only add it selectively on the exception types that use it?

waj · 2020-03-06T17:13:54Z

@straight-shoota actually I like it because it ensures uniformity. Many different new exceptions from shards or std might be created from system errors, and so many errors come from the system. I think we should properly document the from_errno and from_winerror methods and encourage their use. And to be honest, I don't feel that adding a field with a value type is spilling the API. It's convenient and unnoticed most of the time.

jhass · 2020-03-06T18:24:38Z

As a compromise we could also have module SystemError (or SyscallError or so) for it.

waj · 2020-03-06T20:13:11Z

Ok, agreed. I moved the implementation to the SystemError module and included in RuntimeError and IO::Error. Also added some docs to that module 🙂

oprypin · 2020-03-06T22:36:44Z

At a glance, this seems strictly better than #5003
👍

waj · 2020-03-09T13:35:30Z

Thanks @oprypin. I considered some of the ideas and the feedback received on that PR.

… IO::Error

RX14

Apologies for the late (post-merge) review, I have been busy recently...

I really like this change! A lot! This is close to how I always wanted Crystal's exception hierarchy. I have a few small issues we might want to change before releasing though:

RX14 · 2020-03-11T01:16:46Z

+end
+
+class File::Error < IO::Error
+  getter file : String


Would be nice to have some documentation on these.

Of course. Documentation is still missing in many classes and methods.

RX14 · 2020-03-11T01:19:12Z

      errno = Errno.value
      LibC.pthread_sigmask(LibC::SIG_SETMASK, pointerof(oldmask), nil)
-      raise Errno.new("fork", errno)
+      raise RuntimeError.from_errno("fork", errno)


I think it would be kind of nice if Process.new on a non-existing file raised File::NotFoundError, i.e. all file-related errors were handled like this, even in mixed-function syscalls.

RX14 · 2020-03-11T01:29:15Z

+                    EHOSTUNREACH ENOTEMPTY EUSERS EDQUOT ESTALE EREMOTE ENOLCK ENOSYS EOVERFLOW
+                    ECANCELED EIDRM ENOMSG EILSEQ EBADMSG EMULTIHOP ENODATA ENOLINK ENOSR ENOSTR
+                    EPROTO ETIME EOPNOTSUPP ENOTRECOVERABLE EOWNERDEAD) %}
+    {% if LibC.has_constant?(value) %}


We should either move this enum into Crystal:: so it is not public or have it be always the same list of errno despite the platform: this enum must have the same members regardless of platform.

But not all errno values exist in all platforms. Maybe we can map those to EINVAL. But there should be at least a warning if you try to use any of those values in a platform where the value doesn't exist.

We decided a long time ago that the compile-time API must be identical between platforms so that everything compiles on every platform (as long as it uses documented API)

If the errno value doesn't exist on every platform, it just never gets used, never gets raised, so it can be any value.

I don't think we need warnings for this.

The existence of an Error number is one part of the problem. There is platform-specific value for some error codes. For example, grep -R "EDEADLK" src/lib_c/*

src/lib_c/aarch64-linux-gnu/c/errno.cr: EDEADLK = 35 src/lib_c/aarch64-linux-musl/c/errno.cr: EDEADLK = 35 src/lib_c/amd64-unknown-openbsd/c/errno.cr: EDEADLK = 11 src/lib_c/arm-linux-gnueabihf/c/errno.cr: EDEADLK = 35 src/lib_c/i386-linux-gnu/c/errno.cr: EDEADLK = 35 src/lib_c/i386-linux-musl/c/errno.cr: EDEADLK = 35 src/lib_c/i686-linux-gnu/c/errno.cr: EDEADLK = 35 src/lib_c/i686-linux-musl/c/errno.cr: EDEADLK = 35 src/lib_c/x86_64-darwin/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-dragonfly/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-freebsd/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-linux-gnu/c/errno.cr: EDEADLK = 35 src/lib_c/x86_64-linux-musl/c/errno.cr: EDEADLK = 35 src/lib_c/x86_64-macosx-darwin/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-openbsd/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-portbld-freebsd/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-unknown-freebsd/c/errno.cr: EDEADLK = 11 src/lib_c/x86_64-windows-msvc/c/errno.cr: EDEADLK = 36

straight-shoota · 2020-03-11T12:22:33Z

-      io << ")"
-    end
-    raise Errno.new(error_message)
+    raise RuntimeError.from_errno


Why is this not from_error("execvp")?

I think eventually we should make an effort to improve many error messages. I'm not sure it's a good idea to leak the syscall name that failed. Maybe the same error message we have above: "Error executing process"

Yes, that could work. "execvp" has always just been a stand-in for a proper message.

Using an explicit message here would make from_errno without arguments obsolete. I think it should be required to provide at least some message for context information.

Is it leaking the syscall name or telling you exactly what failed, which manpage to read and what to search for?

I actually don't think leaking the name is an issue. It's really pretty concise about defining the context of the error. (Btw. it's a libc call, not a syscall)

But in other cases, I think we can perhaps be more expressive about the error context. There's a huge problem space with localization, though. So IDK, but using just the call looks like a pretty reasonable solution, at least for the time being.

Errno was renamed in 0.34.x crystal-lang/crystal#8885 Instead we should just ignore non existent file. That way it works in 0.33 and 0.34

* Errno was replaced by RuntimeError. See crystal-lang/crystal#8885 * Update Dockerfile to use crystal 0.34.0 * AtExitHandlers was moved to crystal namespace see crystal-lang/crystal#8883 * Use new Log

[Crystal PR #8885](crystal-lang/crystal#8885) changed the name of `IO::Timeout` to `IO::TimeoutError`

The crystal 0.34.0 version includes the removal of Errno and WinError exceptions used by Halite. This last one has released a new version. crystal-lang/crystal#8885 Signed-off-by: Victor Morales <v.morales@samsung.com>

waj added topic:stdlib kind:refactor labels Mar 6, 2020

Fryguy reviewed Mar 6, 2020

View reviewed changes

Comment thread src/exception.cr Outdated

bcardiff reviewed Mar 6, 2020

View reviewed changes

Comment thread src/crystal/system/win32/file.cr Outdated

bcardiff reviewed Mar 6, 2020

View reviewed changes

Comment thread src/socket.cr Outdated

bcardiff approved these changes Mar 6, 2020

View reviewed changes

jhass approved these changes Mar 6, 2020

View reviewed changes

waj force-pushed the feature/no-errno-exceptions branch from 07ef485 to 95484a8 Compare March 9, 2020 13:33

waj added 16 commits March 9, 2020 14:51

Replace usage of Errno with IO::FileSystemError for fs operations

5a42ddc

Replace some untested occurrences of Errno in file system functions

8690d2a

Renamed IO::FileSystemError -> File::Error

80e53b1

Removed exception re-raise at File.size

a541b6d

Refactor File::Error.from_errno into Exception.from_errno

28fdabd

Socket::Error: it now extends IO::Error

4ee40f6

Replaced all occurrences of Errno by Socket::Error in socket classes

f7b28ab

Introduce RuntimeError. Used for iconv errors.

7cdaece

Use RuntimeError within Process class

e2f7195

Replaced all occurrences of Errno exceptions by RuntimeError and some…

8d267ec

… IO::Error

Fixing Windows specs. Raising exceptions from WinError codes.

2feb466

Replaced WinError raised objects by calls to from_winerror

d25f5da

Replaced some remaining rescue Errno by the corresponding class

da47b83

Don't rescue Errno anymore

eb39225

Socket#send raises Socket::ConnectError for ECONNREFUSED

e9bb43f

expect_raises_errno helper not needed anymore

e43e73d

waj added this to the 0.34.0 milestone Mar 10, 2020

waj merged commit 6774221 into crystal-lang:master Mar 10, 2020

waj deleted the feature/no-errno-exceptions branch March 10, 2020 14:54

RX14 reviewed Mar 11, 2020

View reviewed changes

straight-shoota reviewed Mar 11, 2020

View reviewed changes

Sija mentioned this pull request Mar 20, 2020

copy_file_range is a Linux-only optimization for FD to FD copying #8919

Open

j8r mentioned this pull request Mar 24, 2020

Improve Errno error handling #8305

Closed

bcardiff added the breaking-change May have breaking effect in some edge cases. Mostly negligible, but still noteworthy. label Mar 27, 2020

paulcsmith added a commit to luckyframework/lucky_cli that referenced this pull request Apr 5, 2020

Fix Errno issue with Crystal 0.34.0

b8df0ec

Errno was renamed in 0.34.x crystal-lang/crystal#8885 Instead we should just ignore non existent file. That way it works in 0.33 and 0.34

paulcsmith mentioned this pull request Apr 5, 2020

Fix Errno issue with Crystal 0.34.0 and release v0.20.0 luckyframework/lucky_cli#499

Merged

straight-shoota mentioned this pull request Apr 6, 2020

Crystal 0.34.0 will/crystal-pg#203

Merged

paulcsmith added a commit to luckyframework/lucky_cli that referenced this pull request Apr 7, 2020

Fix Errno issue with Crystal 0.34.0

cb94d3c

Errno was renamed in 0.34.x crystal-lang/crystal#8885 Instead we should just ignore non existent file. That way it works in 0.33 and 0.34

anamba mentioned this pull request Apr 8, 2020

Crystal 0.34.0 compatibility jgaskins/neo4j.cr#7

Closed

paulcsmith added a commit to luckyframework/lucky_cli that referenced this pull request Apr 9, 2020

Fix Errno issue with Crystal 0.34.0

457375f

Errno was renamed in 0.34.x crystal-lang/crystal#8885 Instead we should just ignore non existent file. That way it works in 0.33 and 0.34

makenowjust mentioned this pull request Apr 10, 2020

#02 議事録 crystal-jp/weekly#1

Closed

lagerfeuer mentioned this pull request Apr 12, 2020

Errno breaking changes - Crystal 0.34 Papierkorb/fancyline#6

Closed

marcosdsanchez pushed a commit to marcosdsanchez/scry that referenced this pull request Apr 20, 2020

Errno was replaced by RuntimeError. See crystal-lang/crystal#8885

67c7240

marcosdsanchez mentioned this pull request Apr 20, 2020

Support Crystal 0.34.0 crystal-lang-tools/scry#170

Merged

5 tasks

marcosdsanchez pushed a commit to marcosdsanchez/scry that referenced this pull request Apr 20, 2020

Errno was replaced by RuntimeError. See crystal-lang/crystal#8885

c9a7063

This was referenced Jul 4, 2020

Refactor LibC Errnos into separate Enum and add missing Error numbers #9523

Open

Make Errno enum values a set union of all possible LibC Errno values #9573

Draft

dscottboggs added a commit to dscottboggs/cossack that referenced this pull request Jul 8, 2020

Update http_connection.cr

132ba8f

[Crystal PR #8885](crystal-lang/crystal#8885) changed the name of `IO::Timeout` to `IO::TimeoutError`

dscottboggs mentioned this pull request Jul 8, 2020

Update http_connection.cr crystal-community/cossack#22

Draft

This was referenced Dec 3, 2020

add exception filters #1124

Closed

LibC::ESHUTDOWN is not defined #4943

Closed

straight-shoota mentioned this pull request Jun 1, 2021

Abstracting IO timeout settings #10774

Open

Dakad mentioned this pull request Oct 17, 2023

Update http_connection.cr Dakad/cossack.cr#1

Merged

Uh oh!

Conversation

waj commented Mar 6, 2020

Uh oh!

straight-shoota commented Mar 6, 2020

Uh oh!

Uh oh!

waj commented Mar 6, 2020

Uh oh!

jhass commented Mar 6, 2020

Uh oh!

waj commented Mar 6, 2020

Uh oh!

Uh oh!

Uh oh!

oprypin commented Mar 6, 2020

Uh oh!

waj commented Mar 9, 2020

Uh oh!

RX14 left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

RX14 Mar 14, 2020 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

9 participants

RX14 Mar 14, 2020 •

edited

Loading