On Mon, 2007-01-15 at 10:55 +1100, Stephen Rothwell wrote:
> On Sun, 14 Jan 2007 14:54:11 + Alan <[EMAIL PROTECTED]> wrote:
> >
> > This doesn't appea to do the same thing at all. You need to select
> > between two sets of const inode ops instead, otherwise you turn write on
> > arbitarily.
>
On Sun, 14 Jan 2007 14:54:11 + Alan <[EMAIL PROTECTED]> wrote:
>
> This doesn't appea to do the same thing at all. You need to select
> between two sets of const inode ops instead, otherwise you turn write on
> arbitarily.
Or something like below ... (compile tested on pseries, iseries and com
> .read = seq_read,
> + .write = lparcfg_write,
> .open = lparcfg_open,
> .release= single_release,
> };
> @@ -581,10 +582,8 @@ int __init lparcfg_init(void)
>
> /* Allow writing if we have FW_FEATURE_SPLPAR */
> if (firmware
From: Arjan van de Ven <[EMAIL PROTECTED]>
Subject: [patch 00/12] Fix ppc64's writing to struct file_operations
the ppc64 code needlessly wrote to a struct file_operations variable;
this patch turns this into a compile time initialization instead.
Signed-off-by: Arjan van de
4 matches
Mail list logo