Alexander Kuleshov wrote:
> pipe()/pipe2() system calls may fail now with ENOMEM error during creation in
> pipe_create() -> pipespace(). The bellow diff replaces ENOMEM with ENFILE
> errno which is already documented for pipe()/pipe2() to decrease amount of
> possible errno(s) with similar behaviour.
can this error actually occur?
>
> Index: sys_pipe.c
> ===================================================================
> RCS file: /cvs/src/sys/kern/sys_pipe.c,v
> retrieving revision 1.76
> diff -u -p -r1.76 sys_pipe.c
> --- sys_pipe.c 11 Feb 2017 19:51:06 -0000 1.76
> +++ sys_pipe.c 18 Dec 2017 19:19:47 -0000
> @@ -198,7 +198,7 @@ free1:
> * Allocate kva for pipe circular buffer, the space is pageable.
> * This routine will 'realloc' the size of a pipe safely, if it fails
> * it will retain the old buffer.
> - * If it fails it will return ENOMEM.
> + * If it fails it will return ENFILE.
> */
> int
> pipespace(struct pipe *cpipe, u_int size)
> @@ -207,7 +207,7 @@ pipespace(struct pipe *cpipe, u_int size
>
> buffer = km_alloc(size, &kv_any, &kp_pageable, &kd_waitok);
> if (buffer == NULL) {
> - return (ENOMEM);
> + return (ENFILE);
> }
>
> /* free old resources if we are resizing */
>
