xfs
[Top] [All Lists]

Re: [PATCH 3/7] sanitize xlog_in_core_t definition

To: Christoph Hellwig <hch@xxxxxxxxxxxxx>, xfs@xxxxxxxxxxx
Subject: Re: [PATCH 3/7] sanitize xlog_in_core_t definition
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Tue, 28 Oct 2008 05:16:14 -0400
In-reply-to: <20081028051734.GC17077@disturbed>
References: <20081027133907.GD1109@xxxxxxxxxxxxx> <20081028051734.GC17077@disturbed>
User-agent: Mutt/1.5.18 (2008-05-17)
On Tue, Oct 28, 2008 at 04:17:34PM +1100, Dave Chinner wrote:
> > -typedef union xlog_in_core2 {
> > -   xlog_rec_header_t       hic_header;
> > -   xlog_rec_ext_header_t   hic_xheader;
> > -   char                    hic_sector[XLOG_HEADER_SIZE];
> > -} xlog_in_core_2_t;
> > -
> > -typedef struct xlog_in_core {
> > -   xlog_iclog_fields_t     hic_fields;
> > -   xlog_in_core_2_t        *hic_data;
> > +   xlog_in_core_2_t        *ic_data;
> > +#define ic_header  ic_data->hic_header
> >  } xlog_in_core_t;
> 
> The ic_data pointer should not be on the same cacheline as the
> reference count seeing as it is read-only field.

Makes sense, but I'll leave that for a separate patch to keep things
bisectable.

<Prev in Thread] Current Thread [Next in Thread>