Re: [RFC][PATCH 03/12] digest_lists: Basic definitions
From: Greg KH <gregkh@linuxfoundation.org>
Date: 2021-06-27 15:36:04
Also in:
linux-doc, linux-integrity, linux-kselftest, lkml
From: Greg KH <gregkh@linuxfoundation.org>
Date: 2021-06-27 15:36:04
Also in:
linux-doc, linux-integrity, linux-kselftest, lkml
On Sun, Jun 27, 2021 at 04:23:26PM +0100, Matthew Wilcox wrote:
On Sun, Jun 27, 2021 at 12:53:47PM +0200, Greg KH wrote:quoted
quoted
+enum ops { DIGEST_LIST_ADD, DIGEST_LIST_DEL, DIGEST_LIST_OP__LAST }; +For enums you export to userspace, you need to specify the values so that all compilers get them right.I've never heard that rule before. Where does it come from? https://en.cppreference.com/w/c/language/enum says: If enumeration-constant is not followed by = constant-expression, its value is the value one greater than the value of the previous enumerator in the same enumeration. The value of the first enumerator (if it does not use = constant-expression) is zero.
I thought it was in the Documentation/driver-api/ioctl.rst file, but I can't find it right now. Maybe it was something that Arnd said? Arnd, is this still an issue? For some reason I thought it was always good to have uapi .h enums be explicit as to the value in them, otherwise some C compilers might produce other values if they were not specified? thanks, greg k-h