xfs
[Top] [All Lists]

Re: [PATCH 05/11] xfs: optimise xfs_iext_destroy

To: Dave Chinner <david@xxxxxxxxxxxxx>
Subject: Re: [PATCH 05/11] xfs: optimise xfs_iext_destroy
From: Christoph Hellwig <hch@xxxxxxxxxxxxx>
Date: Wed, 13 Apr 2016 09:45:11 -0700
Cc: xfs@xxxxxxxxxxx
Delivered-to: xfs@xxxxxxxxxxx
In-reply-to: <1460525492-1170-6-git-send-email-david@xxxxxxxxxxxxx>
References: <1460525492-1170-1-git-send-email-david@xxxxxxxxxxxxx> <1460525492-1170-6-git-send-email-david@xxxxxxxxxxxxx>
User-agent: Mutt/1.5.24 (2015-08-30)
>  /*
> + * Remove all records from the indirection array.
> + */
> +STATIC void
> +xfs_iext_irec_remove_all(
> +     struct xfs_ifork *ifp)
> +{
> +     int             nlists;
> +     int             i;
> +
> +     ASSERT(ifp->if_flags & XFS_IFEXTIREC);
> +     nlists = ifp->if_real_bytes / XFS_IEXT_BUFSZ;
> +     for (i = 0; i < nlists; i++) {
> +             xfs_ext_irec_t *erp = &ifp->if_u1.if_ext_irec[i];

Can we avoid the typedef for new code?

> +             if (erp->er_extbuf)
> +                     kmem_free(erp->er_extbuf);

no need to check for NULL before a kmem_free, e.g. this whole loop could
just be:

        for (i = 0; i < nlists; i++)
                kmem_free(ifp->if_u1.if_ext_irec[i].er_extbuf);

>  {
>       if (ifp->if_flags & XFS_IFEXTIREC) {
> -             int     erp_idx;
> -             int     nlists;
> -
> -             nlists = ifp->if_real_bytes / XFS_IEXT_BUFSZ;
> -             for (erp_idx = nlists - 1; erp_idx >= 0 ; erp_idx--) {
> -                     xfs_iext_irec_remove(ifp, erp_idx);
> -             }
> +             xfs_iext_irec_remove_all(ifp);
>               ifp->if_flags &= ~XFS_IFEXTIREC;

I'd be tempted to just move clearing of the flag into
xfs_iext_irec_remove_all if we change the patch anyway.

Otherwise this looks fine to me.

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