All of lore.kernel.org
 help / color / mirror / Atom feed
From: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
To: Olof Johansson <olof@lixom.net>
Cc: Dan Williams <djbw@fb.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Vinod Koul <vinod.koul@intel.com>,
	Tomasz Figa <t.figa@samsung.com>,
	Kyungmin Park <kyungmin.park@samsung.com>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Subject: Re: Re: [resend2 PATCH] DMA: remove unused support for MEMSET operations
Date: Fri, 28 Jun 2013 19:15:42 +0200	[thread overview]
Message-ID: <5175148.xKLzH0kDXm@amdc1032> (raw)
In-Reply-To: <CAOesGMgO6ywqk8XzYMdnBfDmB80bu0uPUL2s5kkux-zi5Sa4cw@mail.gmail.com>


Hi,

On Friday, June 28, 2013 09:37:36 AM Olof Johansson wrote:
> Bartlomiej,
> 
> Did you build test on the platforms you changed? You broke dove_defconfig:
> 
> In file included from /home/olof/work/next/include/linux/skbuff.h:31:0,
>                  from /home/olof/work/next/include/linux/if_ether.h:23,
>                  from /home/olof/work/next/include/net/dsa.h:14,
>                  from /home/olof/work/next/arch/arm/plat-orion/common.c:21:
> /home/olof/work/next/arch/arm/plat-orion/common.c: In function
> 'orion_xor1_init':
> /home/olof/work/next/arch/arm/plat-orion/common.c:737:14: error:
> 'DMA_MEMSET' undeclared (first use in this function)
>   dma_cap_set(DMA_MEMSET, orion_xor1_channels_data[1].cap_mask);
>               ^
> /home/olof/work/next/include/linux/dmaengine.h:862:46: note: in
> definition of macro 'dma_cap_set'
>  #define dma_cap_set(tx, mask) __dma_cap_set((tx), &(mask))
>                                               ^
> /home/olof/work/next/arch/arm/plat-orion/common.c:737:14: note: each
> undeclared identifier is reported only once for each function it
> appears in
>   dma_cap_set(DMA_MEMSET, orion_xor1_channels_data[1].cap_mask);
>               ^
> /home/olof/work/next/include/linux/dmaengine.h:862:46: note: in
> definition of macro 'dma_cap_set'
>  #define dma_cap_set(tx, mask) __dma_cap_set((tx), &(mask))
> 
> 
> This would also have been found by a "git grep DMA_MEMSET" across the tree.
> 
> Andrew, this is coming really late before the merge window, I think it
> makes sense to hold off to 3.12 if it can't be done without trivial
> regressions. :(

Sorry for that.  This is an old patch that I ported to current kernels.
I fixed rejects and verified that it compiles for cases that hit rejects
but unfortunately I forgot to check for new code related to DMA_MEMSET
functionality.  Fortunately it seems that the problematic places are in
plat-orion/common.c and mv_xor.c.  They are fixed in the attached patch.

Best regards,
--
Bartlomiej Zolnierkiewicz
Samsung R&D Institute Poland
Samsung Electronics

> -Olof
> 
> On Thu, Jun 27, 2013 at 1:20 PM, Dan Williams <djbw@fb.com> wrote:
> > On Wed, Jun 26, 2013 at 6:31 AM, Bartlomiej Zolnierkiewicz
> > <b.zolnierkie@samsung.com> wrote:
> >> There have never been any real users of MEMSET operations
> >> since they have been introduced in January 2007 (commit
> >> 7405f74badf46b5d023c5d2b670b4471525f6c91 "dmaengine: refactor
> >> dmaengine around dma_async_tx_descriptor").  Therefore remove
> >> support for them for now, it can be always brought back when
> >> needed.
> >>
> >> Cc: Vinod Koul <vinod.koul@intel.com>
> >> Cc: Dan Williams <djbw@fb.com>
> >> Cc: Tomasz Figa <t.figa@samsung.com>
> >> Cc: Andrew Morton <akpm@linux-foundation.org>
> >> Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
> >> Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
> >> ---
> >
> > I see Andrew has already picked this up.
> >
> > Acked-by: Dan Williams <djbw@fb.com>
> > --
> > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> > the body of a message to majordomo@vger.kernel.org
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
> > Please read the FAQ at  http://www.tux.org/lkml/

From: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Subject: [PATCH v2] DMA: remove unused support for MEMSET operations

There have never been any real users of MEMSET operations since they have
been introduced in January 2007 by 7405f74badf4 ("dmaengine: refactor
dmaengine around dma_async_tx_descriptor").  Therefore remove support for
them for now, it can be always brought back when needed.

Acked-by: Dan Williams <djbw@fb.com>
Cc: Vinod Koul <vinod.koul@intel.com>
Cc: Tomasz Figa <t.figa@samsung.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Bartlomiej Zolnierkiewicz <b.zolnierkie@samsung.com>
Signed-off-by: Kyungmin Park <kyungmin.park@samsung.com>
---
v2:
- fixed build problems noticed by Olof
- added Dan's Ack

v1:
- synced against next-20130626

previous submissions:
https://lkml.org/lkml/2012/12/5/116
https://lkml.org/lkml/2012/11/8/107

 Documentation/crypto/async-tx-api.txt |    1 
 arch/arm/mach-iop13xx/setup.c         |    3 
 arch/arm/plat-iop/adma.c              |    2 
 arch/arm/plat-orion/common.c          |   10 --
 crypto/async_tx/Kconfig               |    4 -
 crypto/async_tx/Makefile              |    1 
 crypto/async_tx/async_memset.c        |   89 --------------------------
 drivers/dma/dmaengine.c               |    7 --
 drivers/dma/ioat/dma.c                |    3 
 drivers/dma/ioat/dma_v2.h             |    1 
 drivers/dma/ioat/dma_v3.c             |  114 ----------------------------------
 drivers/dma/ioat/hw.h                 |   27 --------
 drivers/dma/iop-adma.c                |   66 -------------------
 drivers/dma/mv_xor.c                  |   62 ------------------
 drivers/dma/mv_xor.h                  |    1 
 drivers/dma/ppc4xx/adma.c             |   47 --------------
 include/linux/async_tx.h              |    4 -
 include/linux/dmaengine.h             |    5 -
 18 files changed, 7 insertions(+), 440 deletions(-)

Index: b/Documentation/crypto/async-tx-api.txt
===================================================================
--- a/Documentation/crypto/async-tx-api.txt	2013-06-28 19:02:17.237833592 +0200
+++ b/Documentation/crypto/async-tx-api.txt	2013-06-28 19:02:38.945833547 +0200
@@ -222,5 +222,4 @@ drivers/dma/: location for offload engin
 include/linux/async_tx.h: core header file for the async_tx api
 crypto/async_tx/async_tx.c: async_tx interface to dmaengine and common code
 crypto/async_tx/async_memcpy.c: copy offload
-crypto/async_tx/async_memset.c: memory fill offload
 crypto/async_tx/async_xor.c: xor and xor zero sum offload
Index: b/arch/arm/mach-iop13xx/setup.c
===================================================================
--- a/arch/arm/mach-iop13xx/setup.c	2013-06-28 19:02:17.369833591 +0200
+++ b/arch/arm/mach-iop13xx/setup.c	2013-06-28 19:02:38.945833547 +0200
@@ -469,7 +469,6 @@ void __init iop13xx_platform_init(void)
 			dma_cap_set(DMA_MEMCPY, plat_data->cap_mask);
 			dma_cap_set(DMA_XOR, plat_data->cap_mask);
 			dma_cap_set(DMA_XOR_VAL, plat_data->cap_mask);
-			dma_cap_set(DMA_MEMSET, plat_data->cap_mask);
 			dma_cap_set(DMA_INTERRUPT, plat_data->cap_mask);
 			break;
 		case IOP13XX_INIT_ADMA_1:
@@ -479,7 +478,6 @@ void __init iop13xx_platform_init(void)
 			dma_cap_set(DMA_MEMCPY, plat_data->cap_mask);
 			dma_cap_set(DMA_XOR, plat_data->cap_mask);
 			dma_cap_set(DMA_XOR_VAL, plat_data->cap_mask);
-			dma_cap_set(DMA_MEMSET, plat_data->cap_mask);
 			dma_cap_set(DMA_INTERRUPT, plat_data->cap_mask);
 			break;
 		case IOP13XX_INIT_ADMA_2:
@@ -489,7 +487,6 @@ void __init iop13xx_platform_init(void)
 			dma_cap_set(DMA_MEMCPY, plat_data->cap_mask);
 			dma_cap_set(DMA_XOR, plat_data->cap_mask);
 			dma_cap_set(DMA_XOR_VAL, plat_data->cap_mask);
-			dma_cap_set(DMA_MEMSET, plat_data->cap_mask);
 			dma_cap_set(DMA_INTERRUPT, plat_data->cap_mask);
 			dma_cap_set(DMA_PQ, plat_data->cap_mask);
 			dma_cap_set(DMA_PQ_VAL, plat_data->cap_mask);
Index: b/arch/arm/plat-iop/adma.c
===================================================================
--- a/arch/arm/plat-iop/adma.c	2013-06-28 19:02:17.345833592 +0200
+++ b/arch/arm/plat-iop/adma.c	2013-06-28 19:02:38.945833547 +0200
@@ -192,12 +192,10 @@ static int __init iop3xx_adma_cap_init(v
 
 	#ifdef CONFIG_ARCH_IOP32X /* the 32x AAU does not perform zero sum */
 	dma_cap_set(DMA_XOR, iop3xx_aau_data.cap_mask);
-	dma_cap_set(DMA_MEMSET, iop3xx_aau_data.cap_mask);
 	dma_cap_set(DMA_INTERRUPT, iop3xx_aau_data.cap_mask);
 	#else
 	dma_cap_set(DMA_XOR, iop3xx_aau_data.cap_mask);
 	dma_cap_set(DMA_XOR_VAL, iop3xx_aau_data.cap_mask);
-	dma_cap_set(DMA_MEMSET, iop3xx_aau_data.cap_mask);
 	dma_cap_set(DMA_INTERRUPT, iop3xx_aau_data.cap_mask);
 	#endif
 
Index: b/arch/arm/plat-orion/common.c
===================================================================
--- a/arch/arm/plat-orion/common.c	2013-06-28 19:02:17.357833591 +0200
+++ b/arch/arm/plat-orion/common.c	2013-06-28 19:02:38.945833547 +0200
@@ -666,14 +666,9 @@ void __init orion_xor0_init(unsigned lon
 	orion_xor0_shared_resources[3].start = irq_1;
 	orion_xor0_shared_resources[3].end = irq_1;
 
-	/*
-	 * two engines can't do memset simultaneously, this limitation
-	 * satisfied by removing memset support from one of the engines.
-	 */
 	dma_cap_set(DMA_MEMCPY, orion_xor0_channels_data[0].cap_mask);
 	dma_cap_set(DMA_XOR, orion_xor0_channels_data[0].cap_mask);
 
-	dma_cap_set(DMA_MEMSET, orion_xor0_channels_data[1].cap_mask);
 	dma_cap_set(DMA_MEMCPY, orion_xor0_channels_data[1].cap_mask);
 	dma_cap_set(DMA_XOR, orion_xor0_channels_data[1].cap_mask);
 
@@ -732,14 +727,9 @@ void __init orion_xor1_init(unsigned lon
 	orion_xor1_shared_resources[3].start = irq_1;
 	orion_xor1_shared_resources[3].end = irq_1;
 
-	/*
-	 * two engines can't do memset simultaneously, this limitation
-	 * satisfied by removing memset support from one of the engines.
-	 */
 	dma_cap_set(DMA_MEMCPY, orion_xor1_channels_data[0].cap_mask);
 	dma_cap_set(DMA_XOR, orion_xor1_channels_data[0].cap_mask);
 
-	dma_cap_set(DMA_MEMSET, orion_xor1_channels_data[1].cap_mask);
 	dma_cap_set(DMA_MEMCPY, orion_xor1_channels_data[1].cap_mask);
 	dma_cap_set(DMA_XOR, orion_xor1_channels_data[1].cap_mask);
 
Index: b/crypto/async_tx/Kconfig
===================================================================
--- a/crypto/async_tx/Kconfig	2013-06-28 19:02:17.205833592 +0200
+++ b/crypto/async_tx/Kconfig	2013-06-28 19:02:38.945833547 +0200
@@ -10,10 +10,6 @@ config ASYNC_XOR
 	select ASYNC_CORE
 	select XOR_BLOCKS
 
-config ASYNC_MEMSET
-	tristate
-	select ASYNC_CORE
-
 config ASYNC_PQ
 	tristate
 	select ASYNC_CORE
Index: b/crypto/async_tx/Makefile
===================================================================
--- a/crypto/async_tx/Makefile	2013-06-28 19:02:17.225833592 +0200
+++ b/crypto/async_tx/Makefile	2013-06-28 19:02:38.949833547 +0200
@@ -1,6 +1,5 @@
 obj-$(CONFIG_ASYNC_CORE) += async_tx.o
 obj-$(CONFIG_ASYNC_MEMCPY) += async_memcpy.o
-obj-$(CONFIG_ASYNC_MEMSET) += async_memset.o
 obj-$(CONFIG_ASYNC_XOR) += async_xor.o
 obj-$(CONFIG_ASYNC_PQ) += async_pq.o
 obj-$(CONFIG_ASYNC_RAID6_RECOV) += async_raid6_recov.o
Index: b/crypto/async_tx/async_memset.c
===================================================================
--- a/crypto/async_tx/async_memset.c	2013-06-28 19:02:17.213833592 +0200
+++ /dev/null	1970-01-01 00:00:00.000000000 +0000
@@ -1,89 +0,0 @@
-/*
- * memory fill offload engine support
- *
- * Copyright © 2006, Intel Corporation.
- *
- *      Dan Williams <dan.j.williams@intel.com>
- *
- *      with architecture considerations by:
- *      Neil Brown <neilb@suse.de>
- *      Jeff Garzik <jeff@garzik.org>
- *
- * This program is free software; you can redistribute it and/or modify it
- * under the terms and conditions of the GNU General Public License,
- * version 2, as published by the Free Software Foundation.
- *
- * This program is distributed in the hope it will be useful, but WITHOUT
- * ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
- * FITNESS FOR A PARTICULAR PURPOSE.  See the GNU General Public License for
- * more details.
- *
- * You should have received a copy of the GNU General Public License along with
- * this program; if not, write to the Free Software Foundation, Inc.,
- * 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA.
- *
- */
-#include <linux/kernel.h>
-#include <linux/interrupt.h>
-#include <linux/module.h>
-#include <linux/mm.h>
-#include <linux/dma-mapping.h>
-#include <linux/async_tx.h>
-
-/**
- * async_memset - attempt to fill memory with a dma engine.
- * @dest: destination page
- * @val: fill value
- * @offset: offset in pages to start transaction
- * @len: length in bytes
- *
- * honored flags: ASYNC_TX_ACK
- */
-struct dma_async_tx_descriptor *
-async_memset(struct page *dest, int val, unsigned int offset, size_t len,
-	     struct async_submit_ctl *submit)
-{
-	struct dma_chan *chan = async_tx_find_channel(submit, DMA_MEMSET,
-						      &dest, 1, NULL, 0, len);
-	struct dma_device *device = chan ? chan->device : NULL;
-	struct dma_async_tx_descriptor *tx = NULL;
-
-	if (device && is_dma_fill_aligned(device, offset, 0, len)) {
-		dma_addr_t dma_dest;
-		unsigned long dma_prep_flags = 0;
-
-		if (submit->cb_fn)
-			dma_prep_flags |= DMA_PREP_INTERRUPT;
-		if (submit->flags & ASYNC_TX_FENCE)
-			dma_prep_flags |= DMA_PREP_FENCE;
-		dma_dest = dma_map_page(device->dev, dest, offset, len,
-					DMA_FROM_DEVICE);
-
-		tx = device->device_prep_dma_memset(chan, dma_dest, val, len,
-						    dma_prep_flags);
-	}
-
-	if (tx) {
-		pr_debug("%s: (async) len: %zu\n", __func__, len);
-		async_tx_submit(chan, tx, submit);
-	} else { /* run the memset synchronously */
-		void *dest_buf;
-		pr_debug("%s: (sync) len: %zu\n", __func__, len);
-
-		dest_buf = page_address(dest) + offset;
-
-		/* wait for any prerequisite operations */
-		async_tx_quiesce(&submit->depend_tx);
-
-		memset(dest_buf, val, len);
-
-		async_tx_sync_epilog(submit);
-	}
-
-	return tx;
-}
-EXPORT_SYMBOL_GPL(async_memset);
-
-MODULE_AUTHOR("Intel Corporation");
-MODULE_DESCRIPTION("asynchronous memset api");
-MODULE_LICENSE("GPL");
Index: b/drivers/dma/dmaengine.c
===================================================================
--- a/drivers/dma/dmaengine.c	2013-06-28 19:02:17.249833592 +0200
+++ b/drivers/dma/dmaengine.c	2013-06-28 19:02:38.949833547 +0200
@@ -663,11 +663,6 @@ static bool device_has_all_tx_types(stru
 		return false;
 	#endif
 
-	#if defined(CONFIG_ASYNC_MEMSET) || defined(CONFIG_ASYNC_MEMSET_MODULE)
-	if (!dma_has_cap(DMA_MEMSET, device->cap_mask))
-		return false;
-	#endif
-
 	#if defined(CONFIG_ASYNC_XOR) || defined(CONFIG_ASYNC_XOR_MODULE)
 	if (!dma_has_cap(DMA_XOR, device->cap_mask))
 		return false;
@@ -729,8 +724,6 @@ int dma_async_device_register(struct dma
 		!device->device_prep_dma_pq);
 	BUG_ON(dma_has_cap(DMA_PQ_VAL, device->cap_mask) &&
 		!device->device_prep_dma_pq_val);
-	BUG_ON(dma_has_cap(DMA_MEMSET, device->cap_mask) &&
-		!device->device_prep_dma_memset);
 	BUG_ON(dma_has_cap(DMA_INTERRUPT, device->cap_mask) &&
 		!device->device_prep_dma_interrupt);
 	BUG_ON(dma_has_cap(DMA_SG, device->cap_mask) &&
Index: b/drivers/dma/ioat/dma.c
===================================================================
--- a/drivers/dma/ioat/dma.c	2013-06-28 19:02:17.277833592 +0200
+++ b/drivers/dma/ioat/dma.c	2013-06-28 19:02:38.949833547 +0200
@@ -1105,12 +1105,11 @@ static ssize_t cap_show(struct dma_chan
 {
 	struct dma_device *dma = c->device;
 
-	return sprintf(page, "copy%s%s%s%s%s%s\n",
+	return sprintf(page, "copy%s%s%s%s%s\n",
 		       dma_has_cap(DMA_PQ, dma->cap_mask) ? " pq" : "",
 		       dma_has_cap(DMA_PQ_VAL, dma->cap_mask) ? " pq_val" : "",
 		       dma_has_cap(DMA_XOR, dma->cap_mask) ? " xor" : "",
 		       dma_has_cap(DMA_XOR_VAL, dma->cap_mask) ? " xor_val" : "",
-		       dma_has_cap(DMA_MEMSET, dma->cap_mask)  ? " fill" : "",
 		       dma_has_cap(DMA_INTERRUPT, dma->cap_mask) ? " intr" : "");
 
 }
Index: b/drivers/dma/ioat/dma_v2.h
===================================================================
--- a/drivers/dma/ioat/dma_v2.h	2013-06-28 19:02:17.269833592 +0200
+++ b/drivers/dma/ioat/dma_v2.h	2013-06-28 19:02:38.949833547 +0200
@@ -123,7 +123,6 @@ static inline u16 ioat2_xferlen_to_descs
 struct ioat_ring_ent {
 	union {
 		struct ioat_dma_descriptor *hw;
-		struct ioat_fill_descriptor *fill;
 		struct ioat_xor_descriptor *xor;
 		struct ioat_xor_ext_descriptor *xor_ex;
 		struct ioat_pq_descriptor *pq;
Index: b/drivers/dma/ioat/dma_v3.c
===================================================================
--- a/drivers/dma/ioat/dma_v3.c	2013-06-28 19:02:17.289833592 +0200
+++ b/drivers/dma/ioat/dma_v3.c	2013-06-28 19:02:38.949833547 +0200
@@ -311,14 +311,6 @@ static void ioat3_dma_unmap(struct ioat2
 		if (!desc->hw->ctl_f.null) /* skip 'interrupt' ops */
 			ioat_dma_unmap(chan, flags, len, desc->hw);
 		break;
-	case IOAT_OP_FILL: {
-		struct ioat_fill_descriptor *hw = desc->fill;
-
-		if (!(flags & DMA_COMPL_SKIP_DEST_UNMAP))
-			ioat_unmap(pdev, hw->dst_addr - offset, len,
-				   PCI_DMA_FROMDEVICE, flags, 1);
-		break;
-	}
 	case IOAT_OP_XOR_VAL:
 	case IOAT_OP_XOR: {
 		struct ioat_xor_descriptor *xor = desc->xor;
@@ -824,51 +816,6 @@ ioat3_tx_status(struct dma_chan *c, dma_
 }
 
 static struct dma_async_tx_descriptor *
-ioat3_prep_memset_lock(struct dma_chan *c, dma_addr_t dest, int value,
-		       size_t len, unsigned long flags)
-{
-	struct ioat2_dma_chan *ioat = to_ioat2_chan(c);
-	struct ioat_ring_ent *desc;
-	size_t total_len = len;
-	struct ioat_fill_descriptor *fill;
-	u64 src_data = (0x0101010101010101ULL) * (value & 0xff);
-	int num_descs, idx, i;
-
-	num_descs = ioat2_xferlen_to_descs(ioat, len);
-	if (likely(num_descs) && ioat2_check_space_lock(ioat, num_descs) == 0)
-		idx = ioat->head;
-	else
-		return NULL;
-	i = 0;
-	do {
-		size_t xfer_size = min_t(size_t, len, 1 << ioat->xfercap_log);
-
-		desc = ioat2_get_ring_ent(ioat, idx + i);
-		fill = desc->fill;
-
-		fill->size = xfer_size;
-		fill->src_data = src_data;
-		fill->dst_addr = dest;
-		fill->ctl = 0;
-		fill->ctl_f.op = IOAT_OP_FILL;
-
-		len -= xfer_size;
-		dest += xfer_size;
-		dump_desc_dbg(ioat, desc);
-	} while (++i < num_descs);
-
-	desc->txd.flags = flags;
-	desc->len = total_len;
-	fill->ctl_f.int_en = !!(flags & DMA_PREP_INTERRUPT);
-	fill->ctl_f.fence = !!(flags & DMA_PREP_FENCE);
-	fill->ctl_f.compl_write = 1;
-	dump_desc_dbg(ioat, desc);
-
-	/* we leave the channel locked to ensure in order submission */
-	return &desc->txd;
-}
-
-static struct dma_async_tx_descriptor *
 __ioat3_prep_xor_lock(struct dma_chan *c, enum sum_check_flags *result,
 		      dma_addr_t dest, dma_addr_t *src, unsigned int src_cnt,
 		      size_t len, unsigned long flags)
@@ -1431,7 +1378,7 @@ static int ioat_xor_val_self_test(struct
 	struct page *xor_srcs[IOAT_NUM_SRC_TEST];
 	struct page *xor_val_srcs[IOAT_NUM_SRC_TEST + 1];
 	dma_addr_t dma_srcs[IOAT_NUM_SRC_TEST + 1];
-	dma_addr_t dma_addr, dest_dma;
+	dma_addr_t dest_dma;
 	struct dma_async_tx_descriptor *tx;
 	struct dma_chan *dma_chan;
 	dma_cookie_t cookie;
@@ -1598,56 +1545,6 @@ static int ioat_xor_val_self_test(struct
 		goto free_resources;
 	}
 
-	/* skip memset if the capability is not present */
-	if (!dma_has_cap(DMA_MEMSET, dma_chan->device->cap_mask))
-		goto free_resources;
-
-	/* test memset */
-	op = IOAT_OP_FILL;
-
-	dma_addr = dma_map_page(dev, dest, 0,
-			PAGE_SIZE, DMA_FROM_DEVICE);
-	tx = dma->device_prep_dma_memset(dma_chan, dma_addr, 0, PAGE_SIZE,
-					 DMA_PREP_INTERRUPT |
-					 DMA_COMPL_SKIP_SRC_UNMAP |
-					 DMA_COMPL_SKIP_DEST_UNMAP);
-	if (!tx) {
-		dev_err(dev, "Self-test memset prep failed\n");
-		err = -ENODEV;
-		goto dma_unmap;
-	}
-
-	async_tx_ack(tx);
-	init_completion(&cmp);
-	tx->callback = ioat3_dma_test_callback;
-	tx->callback_param = &cmp;
-	cookie = tx->tx_submit(tx);
-	if (cookie < 0) {
-		dev_err(dev, "Self-test memset setup failed\n");
-		err = -ENODEV;
-		goto dma_unmap;
-	}
-	dma->device_issue_pending(dma_chan);
-
-	tmo = wait_for_completion_timeout(&cmp, msecs_to_jiffies(3000));
-
-	if (dma->device_tx_status(dma_chan, cookie, NULL) != DMA_SUCCESS) {
-		dev_err(dev, "Self-test memset timed out\n");
-		err = -ENODEV;
-		goto dma_unmap;
-	}
-
-	dma_unmap_page(dev, dma_addr, PAGE_SIZE, DMA_FROM_DEVICE);
-
-	for (i = 0; i < PAGE_SIZE/sizeof(u32); i++) {
-		u32 *ptr = page_address(dest);
-		if (ptr[i]) {
-			dev_err(dev, "Self-test memset failed compare\n");
-			err = -ENODEV;
-			goto free_resources;
-		}
-	}
-
 	/* test for non-zero parity sum */
 	op = IOAT_OP_XOR_VAL;
 
@@ -1706,8 +1603,7 @@ dma_unmap:
 		for (i = 0; i < IOAT_NUM_SRC_TEST + 1; i++)
 			dma_unmap_page(dev, dma_srcs[i], PAGE_SIZE,
 				       DMA_TO_DEVICE);
-	} else if (op == IOAT_OP_FILL)
-		dma_unmap_page(dev, dma_addr, PAGE_SIZE, DMA_FROM_DEVICE);
+	}
 free_resources:
 	dma->device_free_chan_resources(dma_chan);
 out:
@@ -1944,12 +1840,6 @@ int ioat3_dma_probe(struct ioatdma_devic
 		}
 	}
 
-	if (is_raid_device && (device->cap & IOAT_CAP_FILL_BLOCK)) {
-		dma_cap_set(DMA_MEMSET, dma->cap_mask);
-		dma->device_prep_dma_memset = ioat3_prep_memset_lock;
-	}
-
-
 	dma->device_tx_status = ioat3_tx_status;
 	device->cleanup_fn = ioat3_cleanup_event;
 	device->timer_fn = ioat3_timer_event;
Index: b/drivers/dma/ioat/hw.h
===================================================================
--- a/drivers/dma/ioat/hw.h	2013-06-28 19:02:17.257833592 +0200
+++ b/drivers/dma/ioat/hw.h	2013-06-28 19:02:38.949833547 +0200
@@ -100,33 +100,6 @@ struct ioat_dma_descriptor {
 	uint64_t	user2;
 };
 
-struct ioat_fill_descriptor {
-	uint32_t	size;
-	union {
-		uint32_t ctl;
-		struct {
-			unsigned int int_en:1;
-			unsigned int rsvd:1;
-			unsigned int dest_snoop_dis:1;
-			unsigned int compl_write:1;
-			unsigned int fence:1;
-			unsigned int rsvd2:2;
-			unsigned int dest_brk:1;
-			unsigned int bundle:1;
-			unsigned int rsvd4:15;
-			#define IOAT_OP_FILL 0x01
-			unsigned int op:8;
-		} ctl_f;
-	};
-	uint64_t	src_data;
-	uint64_t	dst_addr;
-	uint64_t	next;
-	uint64_t	rsv1;
-	uint64_t	next_dst_addr;
-	uint64_t	user1;
-	uint64_t	user2;
-};
-
 struct ioat_xor_descriptor {
 	uint32_t	size;
 	union {
Index: b/drivers/dma/iop-adma.c
===================================================================
--- a/drivers/dma/iop-adma.c	2013-06-28 19:02:17.309833591 +0200
+++ b/drivers/dma/iop-adma.c	2013-06-28 19:02:38.949833547 +0200
@@ -633,39 +633,6 @@ iop_adma_prep_dma_memcpy(struct dma_chan
 }
 
 static struct dma_async_tx_descriptor *
-iop_adma_prep_dma_memset(struct dma_chan *chan, dma_addr_t dma_dest,
-			 int value, size_t len, unsigned long flags)
-{
-	struct iop_adma_chan *iop_chan = to_iop_adma_chan(chan);
-	struct iop_adma_desc_slot *sw_desc, *grp_start;
-	int slot_cnt, slots_per_op;
-
-	if (unlikely(!len))
-		return NULL;
-	BUG_ON(len > IOP_ADMA_MAX_BYTE_COUNT);
-
-	dev_dbg(iop_chan->device->common.dev, "%s len: %u\n",
-		__func__, len);
-
-	spin_lock_bh(&iop_chan->lock);
-	slot_cnt = iop_chan_memset_slot_count(len, &slots_per_op);
-	sw_desc = iop_adma_alloc_slots(iop_chan, slot_cnt, slots_per_op);
-	if (sw_desc) {
-		grp_start = sw_desc->group_head;
-		iop_desc_init_memset(grp_start, flags);
-		iop_desc_set_byte_count(grp_start, iop_chan, len);
-		iop_desc_set_block_fill_val(grp_start, value);
-		iop_desc_set_dest_addr(grp_start, iop_chan, dma_dest);
-		sw_desc->unmap_src_cnt = 1;
-		sw_desc->unmap_len = len;
-		sw_desc->async_tx.flags = flags;
-	}
-	spin_unlock_bh(&iop_chan->lock);
-
-	return sw_desc ? &sw_desc->async_tx : NULL;
-}
-
-static struct dma_async_tx_descriptor *
 iop_adma_prep_dma_xor(struct dma_chan *chan, dma_addr_t dma_dest,
 		      dma_addr_t *dma_src, unsigned int src_cnt, size_t len,
 		      unsigned long flags)
@@ -1176,33 +1143,6 @@ iop_adma_xor_val_self_test(struct iop_ad
 		goto free_resources;
 	}
 
-	/* test memset */
-	dma_addr = dma_map_page(dma_chan->device->dev, dest, 0,
-			PAGE_SIZE, DMA_FROM_DEVICE);
-	tx = iop_adma_prep_dma_memset(dma_chan, dma_addr, 0, PAGE_SIZE,
-				      DMA_PREP_INTERRUPT | DMA_CTRL_ACK);
-
-	cookie = iop_adma_tx_submit(tx);
-	iop_adma_issue_pending(dma_chan);
-	msleep(8);
-
-	if (iop_adma_status(dma_chan, cookie, NULL) != DMA_SUCCESS) {
-		dev_err(dma_chan->device->dev,
-			"Self-test memset timed out, disabling\n");
-		err = -ENODEV;
-		goto free_resources;
-	}
-
-	for (i = 0; i < PAGE_SIZE/sizeof(u32); i++) {
-		u32 *ptr = page_address(dest);
-		if (ptr[i]) {
-			dev_err(dma_chan->device->dev,
-				"Self-test memset failed compare, disabling\n");
-			err = -ENODEV;
-			goto free_resources;
-		}
-	}
-
 	/* test for non-zero parity sum */
 	zero_sum_result = 0;
 	for (i = 0; i < IOP_ADMA_NUM_SRC_TEST + 1; i++)
@@ -1487,8 +1427,6 @@ static int iop_adma_probe(struct platfor
 	/* set prep routines based on capability */
 	if (dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask))
 		dma_dev->device_prep_dma_memcpy = iop_adma_prep_dma_memcpy;
-	if (dma_has_cap(DMA_MEMSET, dma_dev->cap_mask))
-		dma_dev->device_prep_dma_memset = iop_adma_prep_dma_memset;
 	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
 		dma_dev->max_xor = iop_adma_get_max_xor();
 		dma_dev->device_prep_dma_xor = iop_adma_prep_dma_xor;
@@ -1556,8 +1494,7 @@ static int iop_adma_probe(struct platfor
 			goto err_free_iop_chan;
 	}
 
-	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask) ||
-	    dma_has_cap(DMA_MEMSET, dma_dev->cap_mask)) {
+	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
 		ret = iop_adma_xor_val_self_test(adev);
 		dev_dbg(&pdev->dev, "xor self test returned %d\n", ret);
 		if (ret)
@@ -1584,7 +1521,6 @@ static int iop_adma_probe(struct platfor
 		 dma_has_cap(DMA_PQ_VAL, dma_dev->cap_mask) ? "pq_val " : "",
 		 dma_has_cap(DMA_XOR, dma_dev->cap_mask) ? "xor " : "",
 		 dma_has_cap(DMA_XOR_VAL, dma_dev->cap_mask) ? "xor_val " : "",
-		 dma_has_cap(DMA_MEMSET, dma_dev->cap_mask)  ? "fill " : "",
 		 dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask) ? "cpy " : "",
 		 dma_has_cap(DMA_INTERRUPT, dma_dev->cap_mask) ? "intr " : "");
 
Index: b/drivers/dma/mv_xor.c
===================================================================
--- a/drivers/dma/mv_xor.c	2013-06-28 19:02:17.333833592 +0200
+++ b/drivers/dma/mv_xor.c	2013-06-28 19:03:01.649833503 +0200
@@ -186,8 +186,6 @@ static int mv_can_chain(struct mv_xor_de
 
 	if (chain_old_tail->type != desc->type)
 		return 0;
-	if (desc->type == DMA_MEMSET)
-		return 0;
 
 	return 1;
 }
@@ -205,9 +203,6 @@ static void mv_set_mode(struct mv_xor_ch
 	case DMA_MEMCPY:
 		op_mode = XOR_OPERATION_MODE_MEMCPY;
 		break;
-	case DMA_MEMSET:
-		op_mode = XOR_OPERATION_MODE_MEMSET;
-		break;
 	default:
 		dev_err(mv_chan_to_devp(chan),
 			"error: unsupported operation %d\n",
@@ -274,18 +269,9 @@ static void mv_xor_start_new_chain(struc
 	if (sw_desc->type != mv_chan->current_type)
 		mv_set_mode(mv_chan, sw_desc->type);
 
-	if (sw_desc->type == DMA_MEMSET) {
-		/* for memset requests we need to program the engine, no
-		 * descriptors used.
-		 */
-		struct mv_xor_desc *hw_desc = sw_desc->hw_desc;
-		mv_chan_set_dest_pointer(mv_chan, hw_desc->phy_dest_addr);
-		mv_chan_set_block_size(mv_chan, sw_desc->unmap_len);
-		mv_chan_set_value(mv_chan, sw_desc->value);
-	} else {
-		/* set the hardware chain */
-		mv_chan_set_next_descriptor(mv_chan, sw_desc->async_tx.phys);
-	}
+	/* set the hardware chain */
+	mv_chan_set_next_descriptor(mv_chan, sw_desc->async_tx.phys);
+
 	mv_chan->pending += sw_desc->slot_cnt;
 	mv_xor_issue_pending(&mv_chan->dmachan);
 }
@@ -688,43 +674,6 @@ mv_xor_prep_dma_memcpy(struct dma_chan *
 }
 
 static struct dma_async_tx_descriptor *
-mv_xor_prep_dma_memset(struct dma_chan *chan, dma_addr_t dest, int value,
-		       size_t len, unsigned long flags)
-{
-	struct mv_xor_chan *mv_chan = to_mv_xor_chan(chan);
-	struct mv_xor_desc_slot *sw_desc, *grp_start;
-	int slot_cnt;
-
-	dev_dbg(mv_chan_to_devp(mv_chan),
-		"%s dest: %x len: %u flags: %ld\n",
-		__func__, dest, len, flags);
-	if (unlikely(len < MV_XOR_MIN_BYTE_COUNT))
-		return NULL;
-
-	BUG_ON(len > MV_XOR_MAX_BYTE_COUNT);
-
-	spin_lock_bh(&mv_chan->lock);
-	slot_cnt = mv_chan_memset_slot_count(len);
-	sw_desc = mv_xor_alloc_slots(mv_chan, slot_cnt, 1);
-	if (sw_desc) {
-		sw_desc->type = DMA_MEMSET;
-		sw_desc->async_tx.flags = flags;
-		grp_start = sw_desc->group_head;
-		mv_desc_init(grp_start, flags);
-		mv_desc_set_byte_count(grp_start, len);
-		mv_desc_set_dest_addr(sw_desc->group_head, dest);
-		mv_desc_set_block_fill_val(grp_start, value);
-		sw_desc->unmap_src_cnt = 1;
-		sw_desc->unmap_len = len;
-	}
-	spin_unlock_bh(&mv_chan->lock);
-	dev_dbg(mv_chan_to_devp(mv_chan),
-		"%s sw_desc %p async_tx %p \n",
-		__func__, sw_desc, &sw_desc->async_tx);
-	return sw_desc ? &sw_desc->async_tx : NULL;
-}
-
-static struct dma_async_tx_descriptor *
 mv_xor_prep_dma_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
 		    unsigned int src_cnt, size_t len, unsigned long flags)
 {
@@ -1137,8 +1086,6 @@ mv_xor_channel_add(struct mv_xor_device
 	/* set prep routines based on capability */
 	if (dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask))
 		dma_dev->device_prep_dma_memcpy = mv_xor_prep_dma_memcpy;
-	if (dma_has_cap(DMA_MEMSET, dma_dev->cap_mask))
-		dma_dev->device_prep_dma_memset = mv_xor_prep_dma_memset;
 	if (dma_has_cap(DMA_XOR, dma_dev->cap_mask)) {
 		dma_dev->max_xor = 8;
 		dma_dev->device_prep_dma_xor = mv_xor_prep_dma_xor;
@@ -1189,7 +1136,6 @@ mv_xor_channel_add(struct mv_xor_device
 
 	dev_info(&pdev->dev, "Marvell XOR: ( %s%s%s%s)\n",
 		 dma_has_cap(DMA_XOR, dma_dev->cap_mask) ? "xor " : "",
-		 dma_has_cap(DMA_MEMSET, dma_dev->cap_mask)  ? "fill " : "",
 		 dma_has_cap(DMA_MEMCPY, dma_dev->cap_mask) ? "cpy " : "",
 		 dma_has_cap(DMA_INTERRUPT, dma_dev->cap_mask) ? "intr " : "");
 
@@ -1298,8 +1244,6 @@ static int mv_xor_probe(struct platform_
 				dma_cap_set(DMA_MEMCPY, cap_mask);
 			if (of_property_read_bool(np, "dmacap,xor"))
 				dma_cap_set(DMA_XOR, cap_mask);
-			if (of_property_read_bool(np, "dmacap,memset"))
-				dma_cap_set(DMA_MEMSET, cap_mask);
 			if (of_property_read_bool(np, "dmacap,interrupt"))
 				dma_cap_set(DMA_INTERRUPT, cap_mask);
 
Index: b/drivers/dma/mv_xor.h
===================================================================
--- a/drivers/dma/mv_xor.h	2013-06-28 19:02:17.321833591 +0200
+++ b/drivers/dma/mv_xor.h	2013-06-28 19:02:38.953833547 +0200
@@ -31,7 +31,6 @@
 
 #define XOR_OPERATION_MODE_XOR		0
 #define XOR_OPERATION_MODE_MEMCPY	2
-#define XOR_OPERATION_MODE_MEMSET	4
 
 #define XOR_CURR_DESC(chan)	(chan->mmr_base + 0x210 + (chan->idx * 4))
 #define XOR_NEXT_DESC(chan)	(chan->mmr_base + 0x200 + (chan->idx * 4))
Index: b/drivers/dma/ppc4xx/adma.c
===================================================================
--- a/drivers/dma/ppc4xx/adma.c	2013-06-28 19:02:17.301833592 +0200
+++ b/drivers/dma/ppc4xx/adma.c	2013-06-28 19:02:38.953833547 +0200
@@ -2323,47 +2323,6 @@ static struct dma_async_tx_descriptor *p
 }
 
 /**
- * ppc440spe_adma_prep_dma_memset - prepare CDB for a MEMSET operation
- */
-static struct dma_async_tx_descriptor *ppc440spe_adma_prep_dma_memset(
-		struct dma_chan *chan, dma_addr_t dma_dest, int value,
-		size_t len, unsigned long flags)
-{
-	struct ppc440spe_adma_chan *ppc440spe_chan;
-	struct ppc440spe_adma_desc_slot *sw_desc, *group_start;
-	int slot_cnt, slots_per_op;
-
-	ppc440spe_chan = to_ppc440spe_adma_chan(chan);
-
-	if (unlikely(!len))
-		return NULL;
-
-	BUG_ON(len > PPC440SPE_ADMA_DMA_MAX_BYTE_COUNT);
-
-	spin_lock_bh(&ppc440spe_chan->lock);
-
-	dev_dbg(ppc440spe_chan->device->common.dev,
-		"ppc440spe adma%d: %s cal: %u len: %u int_en %d\n",
-		ppc440spe_chan->device->id, __func__, value, len,
-		flags & DMA_PREP_INTERRUPT ? 1 : 0);
-
-	slot_cnt = slots_per_op = 1;
-	sw_desc = ppc440spe_adma_alloc_slots(ppc440spe_chan, slot_cnt,
-		slots_per_op);
-	if (sw_desc) {
-		group_start = sw_desc->group_head;
-		ppc440spe_desc_init_memset(group_start, value, flags);
-		ppc440spe_adma_set_dest(group_start, dma_dest, 0);
-		ppc440spe_desc_set_byte_count(group_start, ppc440spe_chan, len);
-		sw_desc->unmap_len = len;
-		sw_desc->async_tx.flags = flags;
-	}
-	spin_unlock_bh(&ppc440spe_chan->lock);
-
-	return sw_desc ? &sw_desc->async_tx : NULL;
-}
-
-/**
  * ppc440spe_adma_prep_dma_xor - prepare CDB for a XOR operation
  */
 static struct dma_async_tx_descriptor *ppc440spe_adma_prep_dma_xor(
@@ -4125,7 +4084,6 @@ static void ppc440spe_adma_init_capabili
 	case PPC440SPE_DMA1_ID:
 		dma_cap_set(DMA_MEMCPY, adev->common.cap_mask);
 		dma_cap_set(DMA_INTERRUPT, adev->common.cap_mask);
-		dma_cap_set(DMA_MEMSET, adev->common.cap_mask);
 		dma_cap_set(DMA_PQ, adev->common.cap_mask);
 		dma_cap_set(DMA_PQ_VAL, adev->common.cap_mask);
 		dma_cap_set(DMA_XOR_VAL, adev->common.cap_mask);
@@ -4151,10 +4109,6 @@ static void ppc440spe_adma_init_capabili
 		adev->common.device_prep_dma_memcpy =
 			ppc440spe_adma_prep_dma_memcpy;
 	}
-	if (dma_has_cap(DMA_MEMSET, adev->common.cap_mask)) {
-		adev->common.device_prep_dma_memset =
-			ppc440spe_adma_prep_dma_memset;
-	}
 	if (dma_has_cap(DMA_XOR, adev->common.cap_mask)) {
 		adev->common.max_xor = XOR_MAX_OPS;
 		adev->common.device_prep_dma_xor =
@@ -4217,7 +4171,6 @@ static void ppc440spe_adma_init_capabili
 	  dma_has_cap(DMA_XOR, adev->common.cap_mask) ? "xor " : "",
 	  dma_has_cap(DMA_XOR_VAL, adev->common.cap_mask) ? "xor_val " : "",
 	  dma_has_cap(DMA_MEMCPY, adev->common.cap_mask) ? "memcpy " : "",
-	  dma_has_cap(DMA_MEMSET, adev->common.cap_mask)  ? "memset " : "",
 	  dma_has_cap(DMA_INTERRUPT, adev->common.cap_mask) ? "intr " : "");
 }
 
Index: b/include/linux/async_tx.h
===================================================================
--- a/include/linux/async_tx.h	2013-06-28 19:02:17.193833592 +0200
+++ b/include/linux/async_tx.h	2013-06-28 19:02:38.953833547 +0200
@@ -182,10 +182,6 @@ async_memcpy(struct page *dest, struct p
 	     unsigned int src_offset, size_t len,
 	     struct async_submit_ctl *submit);
 
-struct dma_async_tx_descriptor *
-async_memset(struct page *dest, int val, unsigned int offset,
-	     size_t len, struct async_submit_ctl *submit);
-
 struct dma_async_tx_descriptor *async_trigger_callback(struct async_submit_ctl *submit);
 
 struct dma_async_tx_descriptor *
Index: b/include/linux/dmaengine.h
===================================================================
--- a/include/linux/dmaengine.h	2013-06-28 19:02:17.181833592 +0200
+++ b/include/linux/dmaengine.h	2013-06-28 19:02:38.953833547 +0200
@@ -66,7 +66,6 @@ enum dma_transaction_type {
 	DMA_PQ,
 	DMA_XOR_VAL,
 	DMA_PQ_VAL,
-	DMA_MEMSET,
 	DMA_INTERRUPT,
 	DMA_SG,
 	DMA_PRIVATE,
@@ -520,7 +519,6 @@ struct dma_tx_state {
  * @device_prep_dma_xor_val: prepares a xor validation operation
  * @device_prep_dma_pq: prepares a pq operation
  * @device_prep_dma_pq_val: prepares a pqzero_sum operation
- * @device_prep_dma_memset: prepares a memset operation
  * @device_prep_dma_interrupt: prepares an end of chain interrupt operation
  * @device_prep_slave_sg: prepares a slave dma operation
  * @device_prep_dma_cyclic: prepare a cyclic dma operation suitable for audio.
@@ -573,9 +571,6 @@ struct dma_device {
 		struct dma_chan *chan, dma_addr_t *pq, dma_addr_t *src,
 		unsigned int src_cnt, const unsigned char *scf, size_t len,
 		enum sum_check_flags *pqres, unsigned long flags);
-	struct dma_async_tx_descriptor *(*device_prep_dma_memset)(
-		struct dma_chan *chan, dma_addr_t dest, int value, size_t len,
-		unsigned long flags);
 	struct dma_async_tx_descriptor *(*device_prep_dma_interrupt)(
 		struct dma_chan *chan, unsigned long flags);
 	struct dma_async_tx_descriptor *(*device_prep_dma_sg)(


  reply	other threads:[~2013-06-28 17:15 UTC|newest]

Thread overview: 5+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-26 13:31 [resend2 PATCH] DMA: remove unused support for MEMSET operations Bartlomiej Zolnierkiewicz
2013-06-27 20:20 ` Dan Williams
2013-06-28 16:37   ` Olof Johansson
2013-06-28 17:15     ` Bartlomiej Zolnierkiewicz [this message]
2013-06-28 22:06       ` Kevin Hilman

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=5175148.xKLzH0kDXm@amdc1032 \
    --to=b.zolnierkie@samsung.com \
    --cc=akpm@linux-foundation.org \
    --cc=djbw@fb.com \
    --cc=kyungmin.park@samsung.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=olof@lixom.net \
    --cc=t.figa@samsung.com \
    --cc=vinod.koul@intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.