"Implement me. Unknown stdin file type!" error thrown when running headless process in win32

[jakelauer]

• Version: 6.10.0
• Platform: Windows 7 x86
• Subsystem:

Hi there - I encountered this while attempting to run PostCSS commands via the PostCSS-CLI command line implementation.

Note - I have only encountered this error in Windows 7. Windows 10 seems to work fine.

The exact error is as shown:

internal/process/stdio.js:82

        throw new Error('Implement me. Unknown stdin file type!');
        ^
    Error: Implement me. Unknown stdin file type!
    at process.getStdin [as stdin] (internal/process/stdio.js:82:15)
    at Object.<anonymous> (C:\Users\myusername\AppData\Local\Microsoft\VisualStudio\14.0\Extensions\5xs3acz0.hfk\Resources\nodejs\node_modules\get-stdin\index.js:2:20)

    at Module._compile (module.js:570:32)
    at Object.Module._extensions..js (module.js:579:10)
    at Module.load (module.js:487:32)
    at tryModuleLoad (module.js:446:12)
    at Function.Module._load (module.js:438:3)
    at Module.require (module.js:497:17)
    at require (internal/module.js:20:19)

    at Object.<anonymous> (C:\Users\myusername\AppData\Local\Microsoft\VisualStudio\14.0\Extensions\5xs3acz0.hfk\Resources\nodejs\node_modules\postcss-cli\index.js:6:15)

I found a similar issue in Angular here: angular/angular-cli#4870

They solved it like this:
angular/angular-cli@4af7a42

Let me know if you need more information - I am very new to Node so you might need to simplify any questions you have for me.

[addaleax]

/cc @nodejs/platform-windows

[seishun]

Can you provide a minimal test case that reproduces the issue, preferably without using third-party libraries?

[addaleax]

The commit message of angular/angular-cli@4af7a42 sounds like a good clue:

When running in a headless process in win32 the lack of process.stdin throws an error.

I can’t test that myself, though.

[seishun]

That doesn't say much. What does "running in a headless process" mean exactly?

[jakelauer]

I'm sorry for the lack of detail - I'm kind of blind to the details of the issue. I'm extra new to Node, so I can't really give you an example without a third-party plugin. I am reproducing the issue while running PostCSS using their postcss-cli module via the command line in Windows, like this:

"cmd /c cd /d [node directory] postcss [arguments]"

Again, apologies for the briefness! I know this is not a lot to go on. I will provide more details if I can as I try to deal with the issue.

[seishun]

In that case I think it would be more appropriate to post the issue in the PostCSS repo.

[jakelauer]

I originally posted it in the repo of the module that actually threw the error (get-stdin): sindresorhus/get-stdin#20

He directed me to post here as NodeJS did not provide code to handle the situation.

[addaleax]

Fwiw, this issue was opened because of sindresorhus/get-stdin#20 (comment) (@jakelauer: Generally, it helps a lot if you could add cross-links when you open issues on multiple repositories!), and it looks to me like this is something that could/should be fixed in Node core.

[jakelauer]

@addaleax Thanks for the pointers :) I haven't had a lot of experience opening issues on GitHub either!

[jakelauer]

Interesting sidenote: I do not experience this in Windows 10

[gireeshpunathil]

I guess 'headless' here means either a non-console application or a windows service. In either case, node is started from a non-shell parent process with custom (or no) stdin. When node is started normally through shell, the C runtime makes sure that the stdin is created and supplied to the process, this is the responsibility of a parent if node is started through createProcess* calls.

The error message indicates that stdin (the fd 0 which node expects stdin to be at the start-up) is invalid, which leads all the suspecion to its creator process.

[addaleax]

Maybe Node should do something like

node/src/node.cc

Lines 4137 to 4147 in cccc6d8

	for (int fd = STDIN_FILENO; fd <= STDERR_FILENO; fd += 1) {
	struct stat ignored;
	if (fstat(fd, &ignored) == 0)
	continue;
	// Anything but EBADF means something is seriously wrong. We don't
	// have to special-case EINTR, fstat() is not interruptible.
	if (errno != EBADF)
	ABORT();
	if (fd != open("/dev/null", O_RDWR))
	ABORT();
	}

for Windows, too? If that makes sense?

[gireeshpunathil]

@addaleax - yes, that makes sense to me - though this will not mitigate (I don't think we should mitigate an invalid descriptor scenario if the parent chose to give us a bad one) the issue, this will make sure that we make the right assertion at the right place.

Let me see if I can come up with a test scenario that fails node in the above manner