Hi Hari,
From: "ext Kanigeri, Hari" <h-kanigeri2@xxxxxx>
Subject: [PATCH] ARM: OMAP add TLB preservation support to IOMMU
Date: Fri, 16 Apr 2010 18:18:59 +0200
> From bcdd232666a163d2661d704f9c21d055bacfd178 Mon Sep 17 00:00:00 2001
> From: Hari Kanigeri <h-kanigeri2@xxxxxx>
> Date: Mon, 8 Mar 2010 18:00:36 -0600
> Subject: [PATCH] ARM: OMAP add TLB preservation support to IOMMU
>
> This patch adds TLB preservation support to IOMMU module
>
> Signed-off-by: Hari Kanigeri <h-kanigeri2@xxxxxx>
> ---
> arch/arm/mach-omap2/iommu2.c | 7 +++++--
> arch/arm/plat-omap/iommu.c | 16 +++++++++-------
> 2 files changed, 14 insertions(+), 9 deletions(-)
>
> diff --git a/arch/arm/mach-omap2/iommu2.c b/arch/arm/mach-omap2/iommu2.c
> index 6f4b7cc..2735bd7 100644
> --- a/arch/arm/mach-omap2/iommu2.c
> +++ b/arch/arm/mach-omap2/iommu2.c
> @@ -146,6 +146,8 @@ static u32 omap2_iommu_fault_isr(struct iommu *obj, u32 *ra)
> printk("\n");
>
> iommu_write_reg(obj, stat, MMU_IRQSTATUS);
> + /* Disable MMU to stop continuous generation of MMU faults */
> + omap2_iommu_disable(obj);
The above comment may look a bit redundant.
> return stat;
> }
>
> @@ -183,7 +185,7 @@ static struct cr_regs *omap2_alloc_cr(struct iommu *obj, struct iotlb_entry *e)
> if (!cr)
> return ERR_PTR(-ENOMEM);
>
> - cr->cam = (e->da & MMU_CAM_VATAG_MASK) | e->prsvd | e->pgsz;
> + cr->cam = (e->da & MMU_CAM_VATAG_MASK) | e->prsvd | e->pgsz | e->valid;
> cr->ram = e->pa | e->endian | e->elsz | e->mixed;
Good finding. This can be a separate patch.
> return cr;
> @@ -211,7 +213,8 @@ static ssize_t omap2_dump_cr(struct iommu *obj, struct cr_regs *cr, char *buf)
> char *p = buf;
>
> /* FIXME: Need more detail analysis of cam/ram */
> - p += sprintf(p, "%08x %08x\n", cr->cam, cr->ram);
> + p += sprintf(p, "%08x %08x %01x\n", cr->cam, cr->ram,
> + (cr->cam & MMU_CAM_P) ? 1 : 0);
This is ok for now.
As described in FIXME comment, this cam/ram whole anaylysis can be
improved/implemented in order to display all items/bits in the end.
> return p - buf;
> }
> diff --git a/arch/arm/plat-omap/iommu.c b/arch/arm/plat-omap/iommu.c
> index 5186728..64d676e 100644
> --- a/arch/arm/plat-omap/iommu.c
> +++ b/arch/arm/plat-omap/iommu.c
> @@ -171,15 +171,12 @@ static void iotlb_lock_get(struct iommu *obj, struct iotlb_lock *l)
> l->base = MMU_LOCK_BASE(val);
> l->vict = MMU_LOCK_VICT(val);
>
> - BUG_ON(l->base != 0); /* Currently no preservation is used */
> }
>
> static void iotlb_lock_set(struct iommu *obj, struct iotlb_lock *l)
> {
> u32 val;
>
> - BUG_ON(l->base != 0); /* Currently no preservation is used */
> -
> val = (l->base << MMU_LOCK_BASE_SHIFT);
> val |= (l->vict << MMU_LOCK_VICT_SHIFT);
>
> @@ -241,7 +238,7 @@ int load_iotlb_entry(struct iommu *obj, struct iotlb_entry *e)
> break;
> }
>
> - if (i == obj->nr_tlb_entries) {
> + if (i == obj->nr_tlb_entries || (l.base == obj->nr_tlb_entries)) {
The above should be dealt separately, since no vacant entry is normal,
but no room to be replaced because of preservation should be warned,
at least.
> dev_dbg(obj->dev, "%s: full: no entry\n", __func__);
> err = -EBUSY;
> goto out;
> @@ -252,13 +249,18 @@ int load_iotlb_entry(struct iommu *obj, struct iotlb_entry *e)
> clk_disable(obj->clk);
> return PTR_ERR(cr);
> }
> -
> iotlb_load_cr(obj, cr);
> kfree(cr);
>
> + /* Increment base number if preservation is set */
> + if (e->prsvd)
> + l.base++;
redundant comment? It tells what the code does.
> /* increment victim for next tlb load */
Although the above is my comment, this looks redundant now too.
> - if (++l.vict == obj->nr_tlb_entries)
> - l.vict = 0;
> + if (++l.vict == obj->nr_tlb_entries) {
> + l.vict = l.base;
> + goto out;
> + }
Does the above work?
> +
> iotlb_lock_set(obj, &l);
> out:
> clk_disable(obj->clk);
> --
> 1.7.0
>
>
> Thank you,
> Best regards,
> Hari
>
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
