Date: Mon, 4 Feb 2013 16:16:29 +1100 (EST) From: Bruce Evans <brde@optusnet.com.au> To: Marius Strobl <marius@FreeBSD.org> Cc: svn-src-head@FreeBSD.org, svn-src-all@FreeBSD.org, src-committers@FreeBSD.org Subject: Re: svn commit: r246301 - head/sys/kern Message-ID: <20130204160720.F1009@besplex.bde.org> In-Reply-To: <201302032143.r13LhuQ3053811@svn.freebsd.org> References: <201302032143.r13LhuQ3053811@svn.freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sun, 3 Feb 2013, Marius Strobl wrote:
> Log:
> Further improve r242655 and supply VM_{MIN,MAX}_KERNEL_ADDRESS as constant
> values to SYSCTL_ULONG(9) where possible.
>
> Submitted by: bde
> Modified: head/sys/kern/kern_malloc.c
> ==============================================================================
> --- head/sys/kern/kern_malloc.c Sun Feb 3 21:30:29 2013 (r246300)
> +++ head/sys/kern/kern_malloc.c Sun Feb 3 21:43:55 2013 (r246301)
> @@ -186,15 +186,16 @@ struct {
> */
> static uma_zone_t mt_zone;
>
> -static vm_offset_t vm_min_kernel_address = VM_MIN_KERNEL_ADDRESS;
> -SYSCTL_ULONG(_vm, OID_AUTO, min_kernel_address, CTLFLAG_RD,
> - &vm_min_kernel_address, 0, "Min kernel address");
> +SYSCTL_ULONG(_vm, OID_AUTO, min_kernel_address, CTLFLAG_RD, NULL,
> + VM_MIN_KERNEL_ADDRESS, "Min kernel address");
SYSCTL_*() (except SYSCTL_PROC()) is conventionally split after
CTLFLAG* (to highlight the usual value pointer). This does a
different splitting, presumably to highlight the value non-pointer.
I'm not sure that that is maintainable. The convention for the
usual case is often not followed.
>
> +SYSCTL_ULONG(_vm, OID_AUTO, max_kernel_address, CTLFLAG_RD,
> #ifndef __sparc64__
> -static vm_offset_t vm_max_kernel_address = VM_MAX_KERNEL_ADDRESS;
> + NULL, VM_MAX_KERNEL_ADDRESS,
> +#else
> + &vm_max_kernel_address, 0,
> #endif
> -SYSCTL_ULONG(_vm, OID_AUTO, max_kernel_address, CTLFLAG_RD,
> - &vm_max_kernel_address, 0, "Max kernel address");
> + "Max kernel address");
I barely remember mentioning that.
MD macros in general could in theory be any expression (say a function
call), but if these ones were non-constant expressions then they
wouldn't have worked in the old code either.
Bruce
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20130204160720.F1009>