#define IOATDMA_H
#include <linux/dmaengine.h>
-#include "hw.h"
-#include "registers.h"
#include <linux/init.h>
#include <linux/dmapool.h>
#include <linux/cache.h>
#include <linux/pci_ids.h>
-#include <net/tcp.h>
+#include <linux/circ_buf.h>
+#include <linux/interrupt.h>
+#include "registers.h"
+#include "hw.h"
#define IOAT_DMA_VERSION "4.00"
-#define IOAT_LOW_COMPLETION_MASK 0xffffffc0
#define IOAT_DMA_DCA_ANY_CPU ~0
-#define to_ioatdma_device(dev) container_of(dev, struct ioatdma_device, common)
-#define to_ioat_desc(lh) container_of(lh, struct ioat_desc_sw, node)
-#define tx_to_ioat_desc(tx) container_of(tx, struct ioat_desc_sw, txd)
-#define to_dev(ioat_chan) (&(ioat_chan)->device->pdev->dev)
-#define to_pdev(ioat_chan) ((ioat_chan)->device->pdev)
+#define to_ioatdma_device(dev) container_of(dev, struct ioatdma_device, dma_dev)
+#define to_dev(ioat_chan) (&(ioat_chan)->ioat_dma->pdev->dev)
+#define to_pdev(ioat_chan) ((ioat_chan)->ioat_dma->pdev)
+
+#define chan_num(ch) ((int)((ch)->reg_base - (ch)->ioat_dma->reg_base) / 0x80)
-#define chan_num(ch) ((int)((ch)->reg_base - (ch)->device->reg_base) / 0x80)
+/* ioat hardware assumes at least two sources for raid operations */
+#define src_cnt_to_sw(x) ((x) + 2)
+#define src_cnt_to_hw(x) ((x) - 2)
+#define ndest_to_sw(x) ((x) + 1)
+#define ndest_to_hw(x) ((x) - 1)
+#define src16_cnt_to_sw(x) ((x) + 9)
+#define src16_cnt_to_hw(x) ((x) - 9)
/*
* workaround for IOAT ver.3.0 null descriptor issue
* @pdev: PCI-Express device
* @reg_base: MMIO register space base address
* @dma_pool: for allocating DMA descriptors
- * @common: embedded struct dma_device
+ * @completion_pool: DMA buffers for completion ops
+ * @sed_hw_pool: DMA super descriptor pools
+ * @dma_dev: embedded struct dma_device
* @version: version of ioatdma device
* @msix_entries: irq handlers
* @idx: per channel data
* @dca: direct cache access context
- * @intr_quirk: interrupt setup quirk (for ioat_v1 devices)
- * @enumerate_channels: hw version specific channel enumeration
- * @reset_hw: hw version specific channel (re)initialization
- * @cleanup_fn: select between the v2 and v3 cleanup routines
- * @timer_fn: select between the v2 and v3 timer watchdog routines
- * @self_test: hardware version specific self test for each supported op type
- *
- * Note: the v3 cleanup routine supports raid operations
+ * @irq_mode: interrupt mode (INTX, MSI, MSIX)
+ * @cap: read DMA capabilities register
*/
struct ioatdma_device {
struct pci_dev *pdev;
struct pci_pool *completion_pool;
#define MAX_SED_POOLS 5
struct dma_pool *sed_hw_pool[MAX_SED_POOLS];
- struct dma_device common;
+ struct dma_device dma_dev;
u8 version;
- struct msix_entry msix_entries[4];
- struct ioat_chan_common *idx[4];
+#define IOAT_MAX_CHANS 4
+ struct msix_entry msix_entries[IOAT_MAX_CHANS];
+ struct ioatdma_chan *idx[IOAT_MAX_CHANS];
struct dca_provider *dca;
enum ioat_irq_mode irq_mode;
u32 cap;
- void (*intr_quirk)(struct ioatdma_device *device);
- int (*enumerate_channels)(struct ioatdma_device *device);
- int (*reset_hw)(struct ioat_chan_common *chan);
- void (*cleanup_fn)(unsigned long data);
- void (*timer_fn)(unsigned long data);
- int (*self_test)(struct ioatdma_device *device);
};
-struct ioat_chan_common {
- struct dma_chan common;
+struct ioatdma_chan {
+ struct dma_chan dma_chan;
void __iomem *reg_base;
dma_addr_t last_completion;
spinlock_t cleanup_lock;
unsigned long state;
- #define IOAT_COMPLETION_PENDING 0
+ #define IOAT_CHAN_DOWN 0
#define IOAT_COMPLETION_ACK 1
#define IOAT_RESET_PENDING 2
#define IOAT_KOBJ_INIT_FAIL 3
#define COMPLETION_TIMEOUT msecs_to_jiffies(100)
#define IDLE_TIMEOUT msecs_to_jiffies(2000)
#define RESET_DELAY msecs_to_jiffies(100)
- struct ioatdma_device *device;
+ struct ioatdma_device *ioat_dma;
dma_addr_t completion_dma;
u64 *completion;
struct tasklet_struct cleanup_task;
struct kobject kobj;
+
+/* ioat v2 / v3 channel attributes
+ * @xfercap_log; log2 of channel max transfer length (for fast division)
+ * @head: allocated index
+ * @issued: hardware notification point
+ * @tail: cleanup index
+ * @dmacount: identical to 'head' except for occasionally resetting to zero
+ * @alloc_order: log2 of the number of allocated descriptors
+ * @produce: number of descriptors to produce at submit time
+ * @ring: software ring buffer implementation of hardware ring
+ * @prep_lock: serializes descriptor preparation (producers)
+ */
+ size_t xfercap_log;
+ u16 head;
+ u16 issued;
+ u16 tail;
+ u16 dmacount;
+ u16 alloc_order;
+ u16 produce;
+ struct ioat_ring_ent **ring;
+ spinlock_t prep_lock;
};
struct ioat_sysfs_entry {
ssize_t (*show)(struct dma_chan *, char *);
};
-/**
- * struct ioat_dma_chan - internal representation of a DMA channel
- */
-struct ioat_dma_chan {
- struct ioat_chan_common base;
-
- size_t xfercap; /* XFERCAP register value expanded out */
-
- spinlock_t desc_lock;
- struct list_head free_desc;
- struct list_head used_desc;
-
- int pending;
- u16 desccount;
- u16 active;
-};
-
/**
* struct ioat_sed_ent - wrapper around super extended hardware descriptor
* @hw: hardware SED
- * @sed_dma: dma address for the SED
- * @list: list member
+ * @dma: dma address for the SED
* @parent: point to the dma descriptor that's the parent
+ * @hw_pool: descriptor pool index
*/
struct ioat_sed_ent {
struct ioat_sed_raw_descriptor *hw;
unsigned int hw_pool;
};
-static inline struct ioat_chan_common *to_chan_common(struct dma_chan *c)
-{
- return container_of(c, struct ioat_chan_common, common);
-}
-
-static inline struct ioat_dma_chan *to_ioat_chan(struct dma_chan *c)
-{
- struct ioat_chan_common *chan = to_chan_common(c);
-
- return container_of(chan, struct ioat_dma_chan, base);
-}
-
-/* wrapper around hardware descriptor format + additional software fields */
-
/**
- * struct ioat_desc_sw - wrapper around hardware descriptor
+ * struct ioat_ring_ent - wrapper around hardware descriptor
* @hw: hardware DMA descriptor (for memcpy)
- * @node: this descriptor will either be on the free list,
- * or attached to a transaction list (tx_list)
+ * @xor: hardware xor descriptor
+ * @xor_ex: hardware xor extension descriptor
+ * @pq: hardware pq descriptor
+ * @pq_ex: hardware pq extension descriptor
+ * @pqu: hardware pq update descriptor
+ * @raw: hardware raw (un-typed) descriptor
* @txd: the generic software descriptor for all engines
+ * @len: total transaction length for unmap
+ * @result: asynchronous result of validate operations
* @id: identifier for debug
+ * @sed: pointer to super extended descriptor sw desc
*/
-struct ioat_desc_sw {
- struct ioat_dma_descriptor *hw;
- struct list_head node;
+
+struct ioat_ring_ent {
+ union {
+ struct ioat_dma_descriptor *hw;
+ struct ioat_xor_descriptor *xor;
+ struct ioat_xor_ext_descriptor *xor_ex;
+ struct ioat_pq_descriptor *pq;
+ struct ioat_pq_ext_descriptor *pq_ex;
+ struct ioat_pq_update_descriptor *pqu;
+ struct ioat_raw_descriptor *raw;
+ };
size_t len;
- struct list_head tx_list;
struct dma_async_tx_descriptor txd;
+ enum sum_check_flags *result;
#ifdef DEBUG
int id;
#endif
+ struct ioat_sed_ent *sed;
};
+extern const struct sysfs_ops ioat_sysfs_ops;
+extern struct ioat_sysfs_entry ioat_version_attr;
+extern struct ioat_sysfs_entry ioat_cap_attr;
+extern int ioat_pending_level;
+extern int ioat_ring_alloc_order;
+extern struct kobj_type ioat_ktype;
+extern struct kmem_cache *ioat_cache;
+extern int ioat_ring_max_alloc_order;
+extern struct kmem_cache *ioat_sed_cache;
+
+static inline struct ioatdma_chan *to_ioat_chan(struct dma_chan *c)
+{
+ return container_of(c, struct ioatdma_chan, dma_chan);
+}
+
+/* wrapper around hardware descriptor format + additional software fields */
#ifdef DEBUG
#define set_desc_id(desc, i) ((desc)->id = (i))
#define desc_id(desc) ((desc)->id)
#endif
static inline void
-__dump_desc_dbg(struct ioat_chan_common *chan, struct ioat_dma_descriptor *hw,
+__dump_desc_dbg(struct ioatdma_chan *ioat_chan, struct ioat_dma_descriptor *hw,
struct dma_async_tx_descriptor *tx, int id)
{
- struct device *dev = to_dev(chan);
+ struct device *dev = to_dev(ioat_chan);
dev_dbg(dev, "desc[%d]: (%#llx->%#llx) cookie: %d flags: %#x"
" ctl: %#10.8x (op: %#x int_en: %d compl: %d)\n", id,
}
#define dump_desc_dbg(c, d) \
- ({ if (d) __dump_desc_dbg(&c->base, d->hw, &d->txd, desc_id(d)); 0; })
+ ({ if (d) __dump_desc_dbg(c, d->hw, &d->txd, desc_id(d)); 0; })
-static inline struct ioat_chan_common *
-ioat_chan_by_index(struct ioatdma_device *device, int index)
+static inline struct ioatdma_chan *
+ioat_chan_by_index(struct ioatdma_device *ioat_dma, int index)
{
- return device->idx[index];
+ return ioat_dma->idx[index];
}
-static inline u64 ioat_chansts_32(struct ioat_chan_common *chan)
+static inline u64 ioat_chansts_32(struct ioatdma_chan *ioat_chan)
{
- u8 ver = chan->device->version;
+ u8 ver = ioat_chan->ioat_dma->version;
u64 status;
u32 status_lo;
/* We need to read the low address first as this causes the
* chipset to latch the upper bits for the subsequent read
*/
- status_lo = readl(chan->reg_base + IOAT_CHANSTS_OFFSET_LOW(ver));
- status = readl(chan->reg_base + IOAT_CHANSTS_OFFSET_HIGH(ver));
+ status_lo = readl(ioat_chan->reg_base + IOAT_CHANSTS_OFFSET_LOW(ver));
+ status = readl(ioat_chan->reg_base + IOAT_CHANSTS_OFFSET_HIGH(ver));
status <<= 32;
status |= status_lo;
#if BITS_PER_LONG == 64
-static inline u64 ioat_chansts(struct ioat_chan_common *chan)
+static inline u64 ioat_chansts(struct ioatdma_chan *ioat_chan)
{
- u8 ver = chan->device->version;
+ u8 ver = ioat_chan->ioat_dma->version;
u64 status;
/* With IOAT v3.3 the status register is 64bit. */
if (ver >= IOAT_VER_3_3)
- status = readq(chan->reg_base + IOAT_CHANSTS_OFFSET(ver));
+ status = readq(ioat_chan->reg_base + IOAT_CHANSTS_OFFSET(ver));
else
- status = ioat_chansts_32(chan);
+ status = ioat_chansts_32(ioat_chan);
return status;
}
#define ioat_chansts ioat_chansts_32
#endif
-static inline void ioat_start(struct ioat_chan_common *chan)
-{
- u8 ver = chan->device->version;
-
- writeb(IOAT_CHANCMD_START, chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
-}
-
static inline u64 ioat_chansts_to_addr(u64 status)
{
return status & IOAT_CHANSTS_COMPLETED_DESCRIPTOR_ADDR;
}
-static inline u32 ioat_chanerr(struct ioat_chan_common *chan)
+static inline u32 ioat_chanerr(struct ioatdma_chan *ioat_chan)
{
- return readl(chan->reg_base + IOAT_CHANERR_OFFSET);
+ return readl(ioat_chan->reg_base + IOAT_CHANERR_OFFSET);
}
-static inline void ioat_suspend(struct ioat_chan_common *chan)
+static inline void ioat_suspend(struct ioatdma_chan *ioat_chan)
{
- u8 ver = chan->device->version;
+ u8 ver = ioat_chan->ioat_dma->version;
- writeb(IOAT_CHANCMD_SUSPEND, chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
+ writeb(IOAT_CHANCMD_SUSPEND,
+ ioat_chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
}
-static inline void ioat_reset(struct ioat_chan_common *chan)
+static inline void ioat_reset(struct ioatdma_chan *ioat_chan)
{
- u8 ver = chan->device->version;
+ u8 ver = ioat_chan->ioat_dma->version;
- writeb(IOAT_CHANCMD_RESET, chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
+ writeb(IOAT_CHANCMD_RESET,
+ ioat_chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
}
-static inline bool ioat_reset_pending(struct ioat_chan_common *chan)
+static inline bool ioat_reset_pending(struct ioatdma_chan *ioat_chan)
{
- u8 ver = chan->device->version;
+ u8 ver = ioat_chan->ioat_dma->version;
u8 cmd;
- cmd = readb(chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
+ cmd = readb(ioat_chan->reg_base + IOAT_CHANCMD_OFFSET(ver));
return (cmd & IOAT_CHANCMD_RESET) == IOAT_CHANCMD_RESET;
}
-static inline void ioat_set_chainaddr(struct ioat_dma_chan *ioat, u64 addr)
-{
- struct ioat_chan_common *chan = &ioat->base;
-
- writel(addr & 0x00000000FFFFFFFF,
- chan->reg_base + IOAT1_CHAINADDR_OFFSET_LOW);
- writel(addr >> 32,
- chan->reg_base + IOAT1_CHAINADDR_OFFSET_HIGH);
-}
-
static inline bool is_ioat_active(unsigned long status)
{
return ((status & IOAT_CHANSTS_STATUS) == IOAT_CHANSTS_ACTIVE);
return !!err;
}
-int ioat_probe(struct ioatdma_device *device);
-int ioat_register(struct ioatdma_device *device);
-int ioat1_dma_probe(struct ioatdma_device *dev, int dca);
-int ioat_dma_self_test(struct ioatdma_device *device);
-void ioat_dma_remove(struct ioatdma_device *device);
+#define IOAT_MAX_ORDER 16
+#define ioat_get_alloc_order() \
+ (min(ioat_ring_alloc_order, IOAT_MAX_ORDER))
+#define ioat_get_max_alloc_order() \
+ (min(ioat_ring_max_alloc_order, IOAT_MAX_ORDER))
+
+static inline u32 ioat_ring_size(struct ioatdma_chan *ioat_chan)
+{
+ return 1 << ioat_chan->alloc_order;
+}
+
+/* count of descriptors in flight with the engine */
+static inline u16 ioat_ring_active(struct ioatdma_chan *ioat_chan)
+{
+ return CIRC_CNT(ioat_chan->head, ioat_chan->tail,
+ ioat_ring_size(ioat_chan));
+}
+
+/* count of descriptors pending submission to hardware */
+static inline u16 ioat_ring_pending(struct ioatdma_chan *ioat_chan)
+{
+ return CIRC_CNT(ioat_chan->head, ioat_chan->issued,
+ ioat_ring_size(ioat_chan));
+}
+
+static inline u32 ioat_ring_space(struct ioatdma_chan *ioat_chan)
+{
+ return ioat_ring_size(ioat_chan) - ioat_ring_active(ioat_chan);
+}
+
+static inline u16
+ioat_xferlen_to_descs(struct ioatdma_chan *ioat_chan, size_t len)
+{
+ u16 num_descs = len >> ioat_chan->xfercap_log;
+
+ num_descs += !!(len & ((1 << ioat_chan->xfercap_log) - 1));
+ return num_descs;
+}
+
+static inline struct ioat_ring_ent *
+ioat_get_ring_ent(struct ioatdma_chan *ioat_chan, u16 idx)
+{
+ return ioat_chan->ring[idx & (ioat_ring_size(ioat_chan) - 1)];
+}
+
+static inline void
+ioat_set_chainaddr(struct ioatdma_chan *ioat_chan, u64 addr)
+{
+ writel(addr & 0x00000000FFFFFFFF,
+ ioat_chan->reg_base + IOAT2_CHAINADDR_OFFSET_LOW);
+ writel(addr >> 32,
+ ioat_chan->reg_base + IOAT2_CHAINADDR_OFFSET_HIGH);
+}
+
+/* IOAT Prep functions */
+struct dma_async_tx_descriptor *
+ioat_dma_prep_memcpy_lock(struct dma_chan *c, dma_addr_t dma_dest,
+ dma_addr_t dma_src, size_t len, unsigned long flags);
+struct dma_async_tx_descriptor *
+ioat_prep_interrupt_lock(struct dma_chan *c, unsigned long flags);
+struct dma_async_tx_descriptor *
+ioat_prep_xor(struct dma_chan *chan, dma_addr_t dest, dma_addr_t *src,
+ unsigned int src_cnt, size_t len, unsigned long flags);
+struct dma_async_tx_descriptor *
+ioat_prep_xor_val(struct dma_chan *chan, dma_addr_t *src,
+ unsigned int src_cnt, size_t len,
+ enum sum_check_flags *result, unsigned long flags);
+struct dma_async_tx_descriptor *
+ioat_prep_pq(struct dma_chan *chan, dma_addr_t *dst, dma_addr_t *src,
+ unsigned int src_cnt, const unsigned char *scf, size_t len,
+ unsigned long flags);
+struct dma_async_tx_descriptor *
+ioat_prep_pq_val(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 *
+ioat_prep_pqxor(struct dma_chan *chan, dma_addr_t dst, dma_addr_t *src,
+ unsigned int src_cnt, size_t len, unsigned long flags);
+struct dma_async_tx_descriptor *
+ioat_prep_pqxor_val(struct dma_chan *chan, dma_addr_t *src,
+ unsigned int src_cnt, size_t len,
+ enum sum_check_flags *result, unsigned long flags);
+
+/* IOAT Operation functions */
+irqreturn_t ioat_dma_do_interrupt(int irq, void *data);
+irqreturn_t ioat_dma_do_interrupt_msix(int irq, void *data);
+struct ioat_ring_ent **
+ioat_alloc_ring(struct dma_chan *c, int order, gfp_t flags);
+void ioat_start_null_desc(struct ioatdma_chan *ioat_chan);
+void ioat_free_ring_ent(struct ioat_ring_ent *desc, struct dma_chan *chan);
+int ioat_reset_hw(struct ioatdma_chan *ioat_chan);
+enum dma_status
+ioat_tx_status(struct dma_chan *c, dma_cookie_t cookie,
+ struct dma_tx_state *txstate);
+void ioat_cleanup_event(unsigned long data);
+void ioat_timer_event(unsigned long data);
+int ioat_check_space_lock(struct ioatdma_chan *ioat_chan, int num_descs);
+void ioat_issue_pending(struct dma_chan *chan);
+void ioat_timer_event(unsigned long data);
+
+/* IOAT Init functions */
+bool is_bwd_ioat(struct pci_dev *pdev);
struct dca_provider *ioat_dca_init(struct pci_dev *pdev, void __iomem *iobase);
-dma_addr_t ioat_get_current_completion(struct ioat_chan_common *chan);
-void ioat_init_channel(struct ioatdma_device *device,
- struct ioat_chan_common *chan, int idx);
-enum dma_status ioat_dma_tx_status(struct dma_chan *c, dma_cookie_t cookie,
- struct dma_tx_state *txstate);
-bool ioat_cleanup_preamble(struct ioat_chan_common *chan,
- dma_addr_t *phys_complete);
-void ioat_kobject_add(struct ioatdma_device *device, struct kobj_type *type);
-void ioat_kobject_del(struct ioatdma_device *device);
-int ioat_dma_setup_interrupts(struct ioatdma_device *device);
-void ioat_stop(struct ioat_chan_common *chan);
-extern const struct sysfs_ops ioat_sysfs_ops;
-extern struct ioat_sysfs_entry ioat_version_attr;
-extern struct ioat_sysfs_entry ioat_cap_attr;
+void ioat_kobject_add(struct ioatdma_device *ioat_dma, struct kobj_type *type);
+void ioat_kobject_del(struct ioatdma_device *ioat_dma);
+int ioat_dma_setup_interrupts(struct ioatdma_device *ioat_dma);
+void ioat_stop(struct ioatdma_chan *ioat_chan);
#endif /* IOATDMA_H */
Code Review / kvmfornfv.git / blobdiff