summaryrefslogtreecommitdiffstats
path: root/arch/arm/mach-davinci
diff options
context:
space:
mode:
authorSudhakar Rajashekhara <sudhakar.raj@ti.com>2009-05-21 07:41:35 -0400
committerKevin Hilman <khilman@deeprootsystems.com>2009-08-26 10:56:56 +0300
commit60902a2cb12c3c1682ee7a04ad7448ec16dc0c29 (patch)
treeba754bff7fadd7106dc9f8549136a514177d0fd1 /arch/arm/mach-davinci
parent4c5adde7943b982d22a7bf711654fbb5cb810667 (diff)
davinci: EDMA: multiple CCs, channel mapping and API changes
- restructure to support multiple channel controllers by using additional struct resources for each CC - interface changes visible to EDMA clients Introduce macros to build IDs from controller and channel number, and to extract them. Modify the edma_alloc_slot function to take an extra argument for the controller. Also update ASoC drivers to use API. ASoC changes Acked-by: Mark Brown <broonie@opensource.wolfsonmicro.com> - Move queue related mappings to dm<soc>.c EDMA in DM355 and DM644x has two transfer controllers while DM646x has four transfer controllers. Moving the queue to tc mapping and queue priority mapping to dm<soc>.c will be helpful to probe these mappings from platform device so that the machine_is_* testing will be avoided. - add channel mapping logic Channel mapping logic is introduced in dm646x EDMA. This implies that there is no fixed association for a channel number to a parameter entry number. In other words, using the DMA channel mapping registers (DCHMAPn), a PaRAM entry can be mapped to any channel. While in the case of dm644x and dm355 there is a fixed mapping between the EDMA channel and Param entry number. Signed-off-by: Naresh Medisetty <naresh@ti.com> Signed-off-by: Sudhakar Rajashekhara <sudhakar.raj@ti.com> Reviewed-by: David Brownell <dbrownell@users.sourceforge.net> Signed-off-by: Kevin Hilman <khilman@deeprootsystems.com>
Diffstat (limited to 'arch/arm/mach-davinci')
-rw-r--r--arch/arm/mach-davinci/devices.c8
-rw-r--r--arch/arm/mach-davinci/dm355.c41
-rw-r--r--arch/arm/mach-davinci/dm644x.c41
-rw-r--r--arch/arm/mach-davinci/dm646x.c46
-rw-r--r--arch/arm/mach-davinci/dma.c811
-rw-r--r--arch/arm/mach-davinci/include/mach/edma.h9
6 files changed, 635 insertions, 321 deletions
diff --git a/arch/arm/mach-davinci/devices.c b/arch/arm/mach-davinci/devices.c
index de16f347566..7a2f8ae89be 100644
--- a/arch/arm/mach-davinci/devices.c
+++ b/arch/arm/mach-davinci/devices.c
@@ -82,10 +82,10 @@ static struct resource mmcsd0_resources[] = {
},
/* DMA channels: RX, then TX */
{
- .start = DAVINCI_DMA_MMCRXEVT,
+ .start = EDMA_CTLR_CHAN(0, DAVINCI_DMA_MMCRXEVT),
.flags = IORESOURCE_DMA,
}, {
- .start = DAVINCI_DMA_MMCTXEVT,
+ .start = EDMA_CTLR_CHAN(0, DAVINCI_DMA_MMCTXEVT),
.flags = IORESOURCE_DMA,
},
};
@@ -119,10 +119,10 @@ static struct resource mmcsd1_resources[] = {
},
/* DMA channels: RX, then TX */
{
- .start = 30, /* rx */
+ .start = EDMA_CTLR_CHAN(0, 30), /* rx */
.flags = IORESOURCE_DMA,
}, {
- .start = 31, /* tx */
+ .start = EDMA_CTLR_CHAN(0, 31), /* tx */
.flags = IORESOURCE_DMA,
},
};
diff --git a/arch/arm/mach-davinci/dm355.c b/arch/arm/mach-davinci/dm355.c
index baaaf328de2..373f0c4003a 100644
--- a/arch/arm/mach-davinci/dm355.c
+++ b/arch/arm/mach-davinci/dm355.c
@@ -558,17 +558,38 @@ static const s8 dma_chan_dm355_no_event[] = {
-1
};
-static struct edma_soc_info dm355_edma_info = {
- .n_channel = 64,
- .n_region = 4,
- .n_slot = 128,
- .n_tc = 2,
- .noevent = dma_chan_dm355_no_event,
+static const s8
+queue_tc_mapping[][2] = {
+ /* {event queue no, TC no} */
+ {0, 0},
+ {1, 1},
+ {-1, -1},
+};
+
+static const s8
+queue_priority_mapping[][2] = {
+ /* {event queue no, Priority} */
+ {0, 3},
+ {1, 7},
+ {-1, -1},
+};
+
+static struct edma_soc_info dm355_edma_info[] = {
+ {
+ .n_channel = 64,
+ .n_region = 4,
+ .n_slot = 128,
+ .n_tc = 2,
+ .n_cc = 1,
+ .noevent = dma_chan_dm355_no_event,
+ .queue_tc_mapping = queue_tc_mapping,
+ .queue_priority_mapping = queue_priority_mapping,
+ },
};
static struct resource edma_resources[] = {
{
- .name = "edma_cc",
+ .name = "edma_cc0",
.start = 0x01c00000,
.end = 0x01c00000 + SZ_64K - 1,
.flags = IORESOURCE_MEM,
@@ -586,10 +607,12 @@ static struct resource edma_resources[] = {
.flags = IORESOURCE_MEM,
},
{
+ .name = "edma0",
.start = IRQ_CCINT0,
.flags = IORESOURCE_IRQ,
},
{
+ .name = "edma0_err",
.start = IRQ_CCERRINT,
.flags = IORESOURCE_IRQ,
},
@@ -598,8 +621,8 @@ static struct resource edma_resources[] = {
static struct platform_device dm355_edma_device = {
.name = "edma",
- .id = -1,
- .dev.platform_data = &dm355_edma_info,
+ .id = 0,
+ .dev.platform_data = dm355_edma_info,
.num_resources = ARRAY_SIZE(edma_resources),
.resource = edma_resources,
};
diff --git a/arch/arm/mach-davinci/dm644x.c b/arch/arm/mach-davinci/dm644x.c
index fb5449b3c97..d145c0cbfe6 100644
--- a/arch/arm/mach-davinci/dm644x.c
+++ b/arch/arm/mach-davinci/dm644x.c
@@ -484,17 +484,38 @@ static const s8 dma_chan_dm644x_no_event[] = {
-1
};
-static struct edma_soc_info dm644x_edma_info = {
- .n_channel = 64,
- .n_region = 4,
- .n_slot = 128,
- .n_tc = 2,
- .noevent = dma_chan_dm644x_no_event,
+static const s8
+queue_tc_mapping[][2] = {
+ /* {event queue no, TC no} */
+ {0, 0},
+ {1, 1},
+ {-1, -1},
+};
+
+static const s8
+queue_priority_mapping[][2] = {
+ /* {event queue no, Priority} */
+ {0, 3},
+ {1, 7},
+ {-1, -1},
+};
+
+static struct edma_soc_info dm644x_edma_info[] = {
+ {
+ .n_channel = 64,
+ .n_region = 4,
+ .n_slot = 128,
+ .n_tc = 2,
+ .n_cc = 1,
+ .noevent = dma_chan_dm644x_no_event,
+ .queue_tc_mapping = queue_tc_mapping,
+ .queue_priority_mapping = queue_priority_mapping,
+ },
};
static struct resource edma_resources[] = {
{
- .name = "edma_cc",
+ .name = "edma_cc0",
.start = 0x01c00000,
.end = 0x01c00000 + SZ_64K - 1,
.flags = IORESOURCE_MEM,
@@ -512,10 +533,12 @@ static struct resource edma_resources[] = {
.flags = IORESOURCE_MEM,
},
{
+ .name = "edma0",
.start = IRQ_CCINT0,
.flags = IORESOURCE_IRQ,
},
{
+ .name = "edma0_err",
.start = IRQ_CCERRINT,
.flags = IORESOURCE_IRQ,
},
@@ -524,8 +547,8 @@ static struct resource edma_resources[] = {
static struct platform_device dm644x_edma_device = {
.name = "edma",
- .id = -1,
- .dev.platform_data = &dm644x_edma_info,
+ .id = 0,
+ .dev.platform_data = dm644x_edma_info,
.num_resources = ARRAY_SIZE(edma_resources),
.resource = edma_resources,
};
diff --git a/arch/arm/mach-davinci/dm646x.c b/arch/arm/mach-davinci/dm646x.c
index 334f0711e0f..e2410738327 100644
--- a/arch/arm/mach-davinci/dm646x.c
+++ b/arch/arm/mach-davinci/dm646x.c
@@ -451,17 +451,43 @@ static const s8 dma_chan_dm646x_no_event[] = {
-1
};
-static struct edma_soc_info dm646x_edma_info = {
- .n_channel = 64,
- .n_region = 6, /* 0-1, 4-7 */
- .n_slot = 512,
- .n_tc = 4,
- .noevent = dma_chan_dm646x_no_event,
+/* Four Transfer Controllers on DM646x */
+static const s8
+dm646x_queue_tc_mapping[][2] = {
+ /* {event queue no, TC no} */
+ {0, 0},
+ {1, 1},
+ {2, 2},
+ {3, 3},
+ {-1, -1},
+};
+
+static const s8
+dm646x_queue_priority_mapping[][2] = {
+ /* {event queue no, Priority} */
+ {0, 4},
+ {1, 0},
+ {2, 5},
+ {3, 1},
+ {-1, -1},
+};
+
+static struct edma_soc_info dm646x_edma_info[] = {
+ {
+ .n_channel = 64,
+ .n_region = 6, /* 0-1, 4-7 */
+ .n_slot = 512,
+ .n_tc = 4,
+ .n_cc = 1,
+ .noevent = dma_chan_dm646x_no_event,
+ .queue_tc_mapping = dm646x_queue_tc_mapping,
+ .queue_priority_mapping = dm646x_queue_priority_mapping,
+ },
};
static struct resource edma_resources[] = {
{
- .name = "edma_cc",
+ .name = "edma_cc0",
.start = 0x01c00000,
.end = 0x01c00000 + SZ_64K - 1,
.flags = IORESOURCE_MEM,
@@ -491,10 +517,12 @@ static struct resource edma_resources[] = {
.flags = IORESOURCE_MEM,
},
{
+ .name = "edma0",
.start = IRQ_CCINT0,
.flags = IORESOURCE_IRQ,
},
{
+ .name = "edma0_err",
.start = IRQ_CCERRINT,
.flags = IORESOURCE_IRQ,
},
@@ -503,8 +531,8 @@ static struct resource edma_resources[] = {
static struct platform_device dm646x_edma_device = {
.name = "edma",
- .id = -1,
- .dev.platform_data = &dm646x_edma_info,
+ .id = 0,
+ .dev.platform_data = dm646x_edma_info,
.num_resources = ARRAY_SIZE(edma_resources),
.resource = edma_resources,
};
diff --git a/arch/arm/mach-davinci/dma.c b/arch/arm/mach-davinci/dma.c
index 15e9eb158bb..5908f7717b2 100644
--- a/arch/arm/mach-davinci/dma.c
+++ b/arch/arm/mach-davinci/dma.c
@@ -100,132 +100,157 @@
#define EDMA_SHADOW0 0x2000 /* 4 regions shadowing global channels */
#define EDMA_PARM 0x4000 /* 128 param entries */
-#define DAVINCI_DMA_3PCC_BASE 0x01C00000
-
#define PARM_OFFSET(param_no) (EDMA_PARM + ((param_no) << 5))
+#define EDMA_DCHMAP 0x0100 /* 64 registers */
+#define CHMAP_EXIST BIT(24)
+
#define EDMA_MAX_DMACH 64
#define EDMA_MAX_PARAMENTRY 512
-#define EDMA_MAX_EVQUE 2 /* FIXME too small */
+#define EDMA_MAX_CC 2
/*****************************************************************************/
-static void __iomem *edmacc_regs_base;
+static void __iomem *edmacc_regs_base[EDMA_MAX_CC];
-static inline unsigned int edma_read(int offset)
+static inline unsigned int edma_read(unsigned ctlr, int offset)
{
- return (unsigned int)__raw_readl(edmacc_regs_base + offset);
+ return (unsigned int)__raw_readl(edmacc_regs_base[ctlr] + offset);
}
-static inline void edma_write(int offset, int val)
+static inline void edma_write(unsigned ctlr, int offset, int val)
{
- __raw_writel(val, edmacc_regs_base + offset);
+ __raw_writel(val, edmacc_regs_base[ctlr] + offset);
}
-static inline void edma_modify(int offset, unsigned and, unsigned or)
+static inline void edma_modify(unsigned ctlr, int offset, unsigned and,
+ unsigned or)
{
- unsigned val = edma_read(offset);
+ unsigned val = edma_read(ctlr, offset);
val &= and;
val |= or;
- edma_write(offset, val);
+ edma_write(ctlr, offset, val);
}
-static inline void edma_and(int offset, unsigned and)
+static inline void edma_and(unsigned ctlr, int offset, unsigned and)
{
- unsigned val = edma_read(offset);
+ unsigned val = edma_read(ctlr, offset);
val &= and;
- edma_write(offset, val);
+ edma_write(ctlr, offset, val);
}
-static inline void edma_or(int offset, unsigned or)
+static inline void edma_or(unsigned ctlr, int offset, unsigned or)
{
- unsigned val = edma_read(offset);
+ unsigned val = edma_read(ctlr, offset);
val |= or;
- edma_write(offset, val);
+ edma_write(ctlr, offset, val);
}
-static inline unsigned int edma_read_array(int offset, int i)
+static inline unsigned int edma_read_array(unsigned ctlr, int offset, int i)
{
- return edma_read(offset + (i << 2));
+ return edma_read(ctlr, offset + (i << 2));
}
-static inline void edma_write_array(int offset, int i, unsigned val)
+static inline void edma_write_array(unsigned ctlr, int offset, int i,
+ unsigned val)
{
- edma_write(offset + (i << 2), val);
+ edma_write(ctlr, offset + (i << 2), val);
}
-static inline void edma_modify_array(int offset, int i,
+static inline void edma_modify_array(unsigned ctlr, int offset, int i,
unsigned and, unsigned or)
{
- edma_modify(offset + (i << 2), and, or);
+ edma_modify(ctlr, offset + (i << 2), and, or);
}
-static inline void edma_or_array(int offset, int i, unsigned or)
+static inline void edma_or_array(unsigned ctlr, int offset, int i, unsigned or)
{
- edma_or(offset + (i << 2), or);
+ edma_or(ctlr, offset + (i << 2), or);
}
-static inline void edma_or_array2(int offset, int i, int j, unsigned or)
+static inline void edma_or_array2(unsigned ctlr, int offset, int i, int j,
+ unsigned or)
{
- edma_or(offset + ((i*2 + j) << 2), or);
+ edma_or(ctlr, offset + ((i*2 + j) << 2), or);
}
-static inline void edma_write_array2(int offset, int i, int j, unsigned val)
+static inline void edma_write_array2(unsigned ctlr, int offset, int i, int j,
+ unsigned val)
{
- edma_write(offset + ((i*2 + j) << 2), val);
+ edma_write(ctlr, offset + ((i*2 + j) << 2), val);
}
-static inline unsigned int edma_shadow0_read(int offset)
+static inline unsigned int edma_shadow0_read(unsigned ctlr, int offset)
{
- return edma_read(EDMA_SHADOW0 + offset);
+ return edma_read(ctlr, EDMA_SHADOW0 + offset);
}
-static inline unsigned int edma_shadow0_read_array(int offset, int i)
+static inline unsigned int edma_shadow0_read_array(unsigned ctlr, int offset,
+ int i)
{
- return edma_read(EDMA_SHADOW0 + offset + (i << 2));
+ return edma_read(ctlr, EDMA_SHADOW0 + offset + (i << 2));
}
-static inline void edma_shadow0_write(int offset, unsigned val)
+static inline void edma_shadow0_write(unsigned ctlr, int offset, unsigned val)
{
- edma_write(EDMA_SHADOW0 + offset, val);
+ edma_write(ctlr, EDMA_SHADOW0 + offset, val);
}
-static inline void edma_shadow0_write_array(int offset, int i, unsigned val)
+static inline void edma_shadow0_write_array(unsigned ctlr, int offset, int i,
+ unsigned val)
{
- edma_write(EDMA_SHADOW0 + offset + (i << 2), val);
+ edma_write(ctlr, EDMA_SHADOW0 + offset + (i << 2), val);
}
-static inline unsigned int edma_parm_read(int offset, int param_no)
+static inline unsigned int edma_parm_read(unsigned ctlr, int offset,
+ int param_no)
{
- return edma_read(EDMA_PARM + offset + (param_no << 5));
+ return edma_read(ctlr, EDMA_PARM + offset + (param_no << 5));
}
-static inline void edma_parm_write(int offset, int param_no, unsigned val)
+static inline void edma_parm_write(unsigned ctlr, int offset, int param_no,
+ unsigned val)
{
- edma_write(EDMA_PARM + offset + (param_no << 5), val);
+ edma_write(ctlr, EDMA_PARM + offset + (param_no << 5), val);
}
-static inline void edma_parm_modify(int offset, int param_no,
+static inline void edma_parm_modify(unsigned ctlr, int offset, int param_no,
unsigned and, unsigned or)
{
- edma_modify(EDMA_PARM + offset + (param_no << 5), and, or);
+ edma_modify(ctlr, EDMA_PARM + offset + (param_no << 5), and, or);
}
-static inline void edma_parm_and(int offset, int param_no, unsigned and)
+static inline void edma_parm_and(unsigned ctlr, int offset, int param_no,
+ unsigned and)
{
- edma_and(EDMA_PARM + offset + (param_no << 5), and);
+ edma_and(ctlr, EDMA_PARM + offset + (param_no << 5), and);
}
-static inline void edma_parm_or(int offset, int param_no, unsigned or)
+static inline void edma_parm_or(unsigned ctlr, int offset, int param_no,
+ unsigned or)
{
- edma_or(EDMA_PARM + offset + (param_no << 5), or);
+ edma_or(ctlr, EDMA_PARM + offset + (param_no << 5), or);
}
/*****************************************************************************/
/* actual number of DMA channels and slots on this silicon */
-static unsigned num_channels;
-static unsigned num_slots;
+struct edma {
+ /* how many dma resources of each type */
+ unsigned num_channels;
+ unsigned num_region;
+ unsigned num_slots;
+ unsigned num_tc;
+ unsigned num_cc;
+
+ /* list of channels with no even trigger; terminated by "-1" */
+ const s8 *noevent;
+
+ /* The edma_inuse bit for each PaRAM slot is clear unless the
+ * channel is in use ... by ARM or DSP, for QDMA, or whatever.
+ */
+ DECLARE_BITMAP(edma_inuse, EDMA_MAX_PARAMENTRY);
-static struct dma_interrupt_data {
- void (*callback)(unsigned channel, unsigned short ch_status,
- void *data);
- void *data;
-} intr_data[EDMA_MAX_DMACH];
+ /* The edma_noevent bit for each channel is clear unless
+ * it doesn't trigger DMA events on this platform. It uses a
+ * bit of SOC-specific initialization code.
+ */
+ DECLARE_BITMAP(edma_noevent, EDMA_MAX_DMACH);
-/* The edma_inuse bit for each PaRAM slot is clear unless the
- * channel is in use ... by ARM or DSP, for QDMA, or whatever.
- */
-static DECLARE_BITMAP(edma_inuse, EDMA_MAX_PARAMENTRY);
+ unsigned irq_res_start;
+ unsigned irq_res_end;
-/* The edma_noevent bit for each channel is clear unless
- * it doesn't trigger DMA events on this platform. It uses a
- * bit of SOC-specific initialization code.
- */
-static DECLARE_BITMAP(edma_noevent, EDMA_MAX_DMACH);
+ struct dma_interrupt_data {
+ void (*callback)(unsigned channel, unsigned short ch_status,
+ void *data);
+ void *data;
+ } intr_data[EDMA_MAX_DMACH];
+};
+
+static struct edma *edma_info[EDMA_MAX_CC];
/* dummy param set used to (re)initialize parameter RAM slots */
static const struct edmacc_param dummy_paramset = {
@@ -233,25 +258,10 @@ static const struct edmacc_param dummy_paramset = {
.ccnt = 1,
};
-static const int __initconst
-queue_tc_mapping[EDMA_MAX_EVQUE + 1][2] = {
-/* {event queue no, TC no} */
- {0, 0},
- {1, 1},
- {-1, -1}
-};
-
-static const int __initconst
-queue_priority_mapping[EDMA_MAX_EVQUE + 1][2] = {
- /* {event queue no, Priority} */
- {0, 3},
- {1, 7},
- {-1, -1}
-};
-
/*****************************************************************************/
-static void map_dmach_queue(unsigned ch_no, enum dma_event_q queue_no)
+static void map_dmach_queue(unsigned ctlr, unsigned ch_no,
+ enum dma_event_q queue_no)
{
int bit = (ch_no & 0x7) * 4;
@@ -260,20 +270,40 @@ static void map_dmach_queue(unsigned ch_no, enum dma_event_q queue_no)
queue_no = EVENTQ_1;
queue_no &= 7;
- edma_modify_array(EDMA_DMAQNUM, (ch_no >> 3),
+ edma_modify_array(ctlr, EDMA_DMAQNUM, (ch_no >> 3),
~(0x7 << bit), queue_no << bit);
}
-static void __init map_queue_tc(int queue_no, int tc_no)
+static void __init map_queue_tc(unsigned ctlr, int queue_no, int tc_no)
{
int bit = queue_no * 4;
- edma_modify(EDMA_QUETCMAP, ~(0x7 << bit), ((tc_no & 0x7) << bit));
+ edma_modify(ctlr, EDMA_QUETCMAP, ~(0x7 << bit), ((tc_no & 0x7) << bit));
}
-static void __init assign_priority_to_queue(int queue_no, int priority)
+static void __init assign_priority_to_queue(unsigned ctlr, int queue_no,
+ int priority)
{
int bit = queue_no * 4;
- edma_modify(EDMA_QUEPRI, ~(0x7 << bit), ((priority & 0x7) << bit));
+ edma_modify(ctlr, EDMA_QUEPRI, ~(0x7 << bit),
+ ((priority & 0x7) << bit));
+}
+
+/**
+ * map_dmach_param - Maps channel number to param entry number
+ *
+ * This maps the dma channel number to param entry numberter. In
+ * other words using the DMA channel mapping registers a param entry
+ * can be mapped to any channel
+ *
+ * Callers are responsible for ensuring the channel mapping logic is
+ * included in that particular EDMA variant (Eg : dm646x)
+ *
+ */
+static void __init map_dmach_param(unsigned ctlr)
+{
+ int i;
+ for (i = 0; i < EDMA_MAX_DMACH; i++)
+ edma_write_array(ctlr, EDMA_DCHMAP , i , (i << 5));
}
static inline void
@@ -281,22 +311,39 @@ setup_dma_interrupt(unsigned lch,
void (*callback)(unsigned channel, u16 ch_status, void *data),
void *data)
{
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(lch);
+ lch = EDMA_CHAN_SLOT(lch);
+
if (!callback) {
- edma_shadow0_write_array(SH_IECR, lch >> 5,
+ edma_shadow0_write_array(ctlr, SH_IECR, lch >> 5,
(1 << (lch & 0x1f)));
}
- intr_data[lch].callback = callback;
- intr_data[lch].data = data;
+ edma_info[ctlr]->intr_data[lch].callback = callback;
+ edma_info[ctlr]->intr_data[lch].data = data;
if (callback) {
- edma_shadow0_write_array(SH_ICR, lch >> 5,
+ edma_shadow0_write_array(ctlr, SH_ICR, lch >> 5,
(1 << (lch & 0x1f)));
- edma_shadow0_write_array(SH_IESR, lch >> 5,
+ edma_shadow0_write_array(ctlr, SH_IESR, lch >> 5,
(1 << (lch & 0x1f)));
}
}
+static int irq2ctlr(int irq)
+{
+ if (irq >= edma_info[0]->irq_res_start &&
+ irq <= edma_info[0]->irq_res_end)
+ return 0;
+ else if (irq >= edma_info[1]->irq_res_start &&
+ irq <= edma_info[1]->irq_res_end)
+ return 1;
+
+ return -1;
+}
+
/******************************************************************************
*
* DMA interrupt handler
@@ -305,32 +352,39 @@ setup_dma_interrupt(unsigned lch,
static irqreturn_t dma_irq_handler(int irq, void *data)
{
int i;
+ unsigned ctlr;
unsigned int cnt = 0;
+ ctlr = irq2ctlr(irq);
+
dev_dbg(data, "dma_irq_handler\n");
- if ((edma_shadow0_read_array(SH_IPR, 0) == 0)
- && (edma_shadow0_read_array(SH_IPR, 1) == 0))
+ if ((edma_shadow0_read_array(ctlr, SH_IPR, 0) == 0)
+ && (edma_shadow0_read_array(ctlr, SH_IPR, 1) == 0))
return IRQ_NONE;
while (1) {
int j;
- if (edma_shadow0_read_array(SH_IPR, 0))
+ if (edma_shadow0_read_array(ctlr, SH_IPR, 0))
j = 0;
- else if (edma_shadow0_read_array(SH_IPR, 1))
+ else if (edma_shadow0_read_array(ctlr, SH_IPR, 1))
j = 1;
else
break;
dev_dbg(data, "IPR%d %08x\n", j,
- edma_shadow0_read_array(SH_IPR, j));
+ edma_shadow0_read_array(ctlr, SH_IPR, j));
for (i = 0; i < 32; i++) {
int k = (j << 5) + i;
- if (edma_shadow0_read_array(SH_IPR, j) & (1 << i)) {
+ if (edma_shadow0_read_array(ctlr, SH_IPR, j) &
+ (1 << i)) {
/* Clear the corresponding IPR bits */
- edma_shadow0_write_array(SH_ICR, j, (1 << i));
- if (intr_data[k].callback) {
- intr_data[k].callback(k, DMA_COMPLETE,
- intr_data[k].data);
+ edma_shadow0_write_array(ctlr, SH_ICR, j,
+ (1 << i));
+ if (edma_info[ctlr]->intr_data[k].callback) {
+ edma_info[ctlr]->intr_data[k].callback(
+ k, DMA_COMPLETE,
+ edma_info[ctlr]->intr_data[k].
+ data);
}
}
}
@@ -338,7 +392,7 @@ static irqreturn_t dma_irq_handler(int irq, void *data)
if (cnt > 10)
break;
}
- edma_shadow0_write(SH_IEVAL, 1);
+ edma_shadow0_write(ctlr, SH_IEVAL, 1);
return IRQ_HANDLED;
}
@@ -350,78 +404,87 @@ static irqreturn_t dma_irq_handler(int irq, void *data)
static irqreturn_t dma_ccerr_handler(int irq, void *data)
{
int i;
+ unsigned ctlr;
unsigned int cnt = 0;
+ ctlr = irq2ctlr(irq);
+
dev_dbg(data, "dma_ccerr_handler\n");
- if ((edma_read_array(EDMA_EMR, 0) == 0) &&
- (edma_read_array(EDMA_EMR, 1) == 0) &&
- (edma_read(EDMA_QEMR) == 0) && (edma_read(EDMA_CCERR) == 0))
+ if ((edma_read_array(ctlr, EDMA_EMR, 0) == 0) &&
+ (edma_read_array(ctlr, EDMA_EMR, 1) == 0) &&
+ (edma_read(ctlr, EDMA_QEMR) == 0) &&
+ (edma_read(ctlr, EDMA_CCERR) == 0))
return IRQ_NONE;
while (1) {
int j = -1;
- if (edma_read_array(EDMA_EMR, 0))
+ if (edma_read_array(ctlr, EDMA_EMR, 0))
j = 0;
- else if (edma_read_array(EDMA_EMR, 1))
+ else if (edma_read_array(ctlr, EDMA_EMR, 1))
j = 1;
if (j >= 0) {
dev_dbg(data, "EMR%d %08x\n", j,
- edma_read_array(EDMA_EMR, j));
+ edma_read_array(ctlr, EDMA_EMR, j));
for (i = 0; i < 32; i++) {
int k = (j << 5) + i;
- if (edma_read_array(EDMA_EMR, j) & (1 << i)) {
+ if (edma_read_array(ctlr, EDMA_EMR, j) &
+ (1 << i)) {
/* Clear the corresponding EMR bits */
- edma_write_array(EDMA_EMCR, j, 1 << i);
+ edma_write_array(ctlr, EDMA_EMCR, j,
+ 1 << i);
/* Clear any SER */
- edma_shadow0_write_array(SH_SECR, j,
- (1 << i));
- if (intr_data[k].callback) {
- intr_data[k].callback(k,
- DMA_CC_ERROR,
- intr_data
- [k].data);
+ edma_shadow0_write_array(ctlr, SH_SECR,
+ j, (1 << i));
+ if (edma_info[ctlr]->intr_data[k].
+ callback) {
+ edma_info[ctlr]->intr_data[k].
+ callback(k,
+ DMA_CC_ERROR,
+ edma_info[ctlr]->intr_data
+ [k].data);
}
}
}
- } else if (edma_read(EDMA_QEMR)) {
+ } else if (edma_read(ctlr, EDMA_QEMR)) {
dev_dbg(data, "QEMR %02x\n",
- edma_read(EDMA_QEMR));
+ edma_read(ctlr, EDMA_QEMR));
for (i = 0; i < 8; i++) {
- if (edma_read(EDMA_QEMR) & (1 << i)) {
+ if (edma_read(ctlr, EDMA_QEMR) & (1 << i)) {
/* Clear the corresponding IPR bits */
- edma_write(EDMA_QEMCR, 1 << i);
- edma_shadow0_write(SH_QSECR, (1 << i));
+ edma_write(ctlr, EDMA_QEMCR, 1 << i);
+ edma_shadow0_write(ctlr, SH_QSECR,
+ (1 << i));
/* NOTE: not reported!! */
}
}
- } else if (edma_read(EDMA_CCERR)) {
+ } else if (edma_read(ctlr, EDMA_CCERR)) {
dev_dbg(data, "CCERR %08x\n",
- edma_read(EDMA_CCERR));
+ edma_read(ctlr, EDMA_CCERR));
/* FIXME: CCERR.BIT(16) ignored! much better
* to just write CCERRCLR with CCERR value...
*/
for (i = 0; i < 8; i++) {
- if (edma_read(EDMA_CCERR) & (1 << i)) {
+ if (edma_read(ctlr, EDMA_CCERR) & (1 << i)) {
/* Clear the corresponding IPR bits */
- edma_write(EDMA_CCERRCLR, 1 << i);
+ edma_write(ctlr, EDMA_CCERRCLR, 1 << i);
/* NOTE: not reported!! */
}
}
}
- if ((edma_read_array(EDMA_EMR, 0) == 0)
- && (edma_read_array(EDMA_EMR, 1) == 0)
- && (edma_read(EDMA_QEMR) == 0)
- && (edma_read(EDMA_CCERR) == 0)) {
+ if ((edma_read_array(ctlr, EDMA_EMR, 0) == 0)
+ && (edma_read_array(ctlr, EDMA_EMR, 1) == 0)
+ && (edma_read(ctlr, EDMA_QEMR) == 0)
+ && (edma_read(ctlr, EDMA_CCERR) == 0)) {
break;
}
cnt++;
if (cnt > 10)
break;
}
- edma_write(EDMA_EEVAL, 1);
+ edma_write(ctlr, EDMA_EEVAL, 1);
return IRQ_HANDLED;
}
@@ -484,35 +547,53 @@ int edma_alloc_channel(int channel,
void *data,
enum dma_event_q eventq_no)
{
+ unsigned i, done, ctlr = 0;
+
+ if (channel >= 0) {
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+ }
+
if (channel < 0) {
- channel = 0;
- for (;;) {
- channel = find_next_bit(edma_noevent,
- num_channels, channel);
- if (channel == num_channels)
- return -ENOMEM;
- if (!test_and_set_bit(channel, edma_inuse))
+ for (i = 0; i < EDMA_MAX_CC; i++) {
+ channel = 0;
+ for (;;) {
+ channel = find_next_bit(edma_info[i]->
+ edma_noevent,
+ edma_info[i]->num_channels,
+ channel);
+ if (channel == edma_info[i]->num_channels)
+ return -ENOMEM;
+ if (!test_and_set_bit(channel,
+ edma_info[i]->edma_inuse)) {
+ done = 1;
+ ctlr = i;
+ break;
+ }
+ channel++;
+ }
+ if (done)
break;
- channel++;
}
- } else if (channel >= num_channels) {
+ } else if (channel >= edma_info[ctlr]->num_channels) {
return -EINVAL;
- } else if (test_and_set_bit(channel, edma_inuse)) {
+ } else if (test_and_set_bit(channel, edma_info[ctlr]->edma_inuse)) {
return -EBUSY;
}
/* ensure access through shadow region 0 */
- edma_or_array2(EDMA_DRAE, 0, channel >> 5, 1 << (channel & 0x1f));
+ edma_or_array2(ctlr, EDMA_DRAE, 0, channel >> 5, 1 << (channel & 0x1f));
/* ensure no events are pending */
- edma_stop(channel);
- memcpy_toio(edmacc_regs_base + PARM_OFFSET(channel),
+ edma_stop(EDMA_CTLR_CHAN(ctlr, channel));
+ memcpy_toio(edmacc_regs_base[ctlr] + PARM_OFFSET(channel),
&dummy_paramset, PARM_SIZE);
if (callback)
- setup_dma_interrupt(channel, callback, data);
+ setup_dma_interrupt(EDMA_CTLR_CHAN(ctlr, channel),
+ callback, data);
- map_dmach_queue(channel, eventq_no);
+ map_dmach_queue(ctlr, channel, eventq_no);
return channel;
}
@@ -532,15 +613,20 @@ EXPORT_SYMBOL(edma_alloc_channel);
*/
void edma_free_channel(unsigned channel)
{
- if (channel >= num_channels)
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel >= edma_info[ctlr]->num_channels)
return;
setup_dma_interrupt(channel, NULL, NULL);
/* REVISIT should probably take out of shadow region 0 */
- memcpy_toio(edmacc_regs_base + PARM_OFFSET(channel),
+ memcpy_toio(edmacc_regs_base[ctlr] + PARM_OFFSET(channel),
&dummy_paramset, PARM_SIZE);
- clear_bit(channel, edma_inuse);
+ clear_bit(channel, edma_info[ctlr]->edma_inuse);
}
EXPORT_SYMBOL(edma_free_channel);
@@ -558,28 +644,33 @@ EXPORT_SYMBOL(edma_free_channel);
*
* Returns the number of the slot, else negative errno.
*/
-int edma_alloc_slot(int slot)
+int edma_alloc_slot(unsigned ctlr, int slot)
{
+ if (slot >= 0)
+ slot = EDMA_CHAN_SLOT(slot);
+
if (slot < 0) {
- slot = num_channels;
+ slot = edma_info[ctlr]->num_channels;
for (;;) {
- slot = find_next_zero_bit(edma_inuse,
- num_slots, slot);
- if (slot == num_slots)
+ slot = find_next_zero_bit(edma_info[ctlr]->edma_inuse,
+ edma_info[ctlr]->num_slots, slot);
+ if (slot == edma_info[ctlr]->num_slots)
return -ENOMEM;
- if (!test_and_set_bit(slot, edma_inuse))
+ if (!test_and_set_bit(slot,
+ edma_info[ctlr]->edma_inuse))
break;
}
- } else if (slot < num_channels || slot >= num_slots) {
+ } else if (slot < edma_info[ctlr]->num_channels ||
+ slot >= edma_info[ctlr]->num_slots) {
return -EINVAL;
- } else if (test_and_set_bit(slot, edma_inuse)) {
+ } else if (test_and_set_bit(slot, edma_info[ctlr]->edma_inuse)) {
return -EBUSY;
}
- memcpy_toio(edmacc_regs_base + PARM_OFFSET(slot),
+ memcpy_toio(edmacc_regs_base[ctlr] + PARM_OFFSET(slot),
&dummy_paramset, PARM_SIZE);
- return slot;
+ return EDMA_CTLR_CHAN(ctlr, slot);
}
EXPORT_SYMBOL(edma_alloc_slot);
@@ -593,12 +684,18 @@ EXPORT_SYMBOL(edma_alloc_slot);
*/
void edma_free_slot(unsigned slot)
{
- if (slot < num_channels || slot >= num_slots)
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot < edma_info[ctlr]->num_channels ||
+ slot >= edma_info[ctlr]->num_slots)
return;
- memcpy_toio(edmacc_regs_base + PARM_OFFSET(slot),
+ memcpy_toio(edmacc_regs_base[ctlr] + PARM_OFFSET(slot),
&dummy_paramset, PARM_SIZE);
- clear_bit(slot, edma_inuse);
+ clear_bit(slot, edma_info[ctlr]->edma_inuse);
}
EXPORT_SYMBOL(edma_free_slot);
@@ -620,8 +717,13 @@ EXPORT_SYMBOL(edma_free_slot);
void edma_set_src(unsigned slot, dma_addr_t src_port,
enum address_mode mode, enum fifo_width width)
{
- if (slot < num_slots) {
- unsigned int i = edma_parm_read(PARM_OPT, slot);
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot < edma_info[ctlr]->num_slots) {
+ unsigned int i = edma_parm_read(ctlr, PARM_OPT, slot);
if (mode) {
/* set SAM and program FWID */
@@ -630,11 +732,11 @@ void edma_set_src(unsigned slot, dma_addr_t src_port,
/* clear SAM */
i &= ~SAM;
}
- edma_parm_write(PARM_OPT, slot, i);
+ edma_parm_write(ctlr, PARM_OPT, slot, i);
/* set the source port address
in source register of param structure */
- edma_parm_write(PARM_SRC, slot, src_port);
+ edma_parm_write(ctlr, PARM_SRC, slot, src_port);
}
}
EXPORT_SYMBOL(edma_set_src);
@@ -653,8 +755,13 @@ EXPORT_SYMBOL(edma_set_src);
void edma_set_dest(unsigned slot, dma_addr_t dest_port,
enum address_mode mode, enum fifo_width width)
{
- if (slot < num_slots) {
- unsigned int i = edma_parm_read(PARM_OPT, slot);
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot < edma_info[ctlr]->num_slots) {
+ unsigned int i = edma_parm_read(ctlr, PARM_OPT, slot);
if (mode) {
/* set DAM and program FWID */
@@ -663,10 +770,10 @@ void edma_set_dest(unsigned slot, dma_addr_t dest_port,
/* clear DAM */
i &= ~DAM;
}
- edma_parm_write(PARM_OPT, slot, i);
+ edma_parm_write(ctlr, PARM_OPT, slot, i);
/* set the destination port address
in dest register of param structure */
- edma_parm_write(PARM_DST, slot, dest_port);
+ edma_parm_write(ctlr, PARM_DST, slot, dest_port);
}
}
EXPORT_SYMBOL(edma_set_dest);
@@ -683,8 +790,12 @@ EXPORT_SYMBOL(edma_set_dest);
void edma_get_position(unsigned slot, dma_addr_t *src, dma_addr_t *dst)
{
struct edmacc_param temp;
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
- edma_read_slot(slot, &temp);
+ edma_read_slot(EDMA_CTLR_CHAN(ctlr, slot), &temp);
if (src != NULL)
*src = temp.src;
if (dst != NULL)
@@ -704,10 +815,15 @@ EXPORT_SYMBOL(edma_get_position);
*/
void edma_set_src_index(unsigned slot, s16 src_bidx, s16 src_cidx)
{
- if (slot < num_slots) {
- edma_parm_modify(PARM_SRC_DST_BIDX, slot,
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot < edma_info[ctlr]->num_slots) {
+ edma_parm_modify(ctlr, PARM_SRC_DST_BIDX, slot,
0xffff0000, src_bidx);
- edma_parm_modify(PARM_SRC_DST_CIDX, slot,
+ edma_parm_modify(ctlr, PARM_SRC_DST_CIDX, slot,
0xffff0000, src_cidx);
}
}
@@ -725,10 +841,15 @@ EXPORT_SYMBOL(edma_set_src_index);
*/
void edma_set_dest_index(unsigned slot, s16 dest_bidx, s16 dest_cidx)
{
- if (slot < num_slots) {
- edma_parm_modify(PARM_SRC_DST_BIDX, slot,
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot < edma_info[ctlr]->num_slots) {
+ edma_parm_modify(ctlr, PARM_SRC_DST_BIDX, slot,
0x0000ffff, dest_bidx << 16);
- edma_parm_modify(PARM_SRC_DST_CIDX, slot,
+ edma_parm_modify(ctlr, PARM_SRC_DST_CIDX, slot,
0x0000ffff, dest_cidx << 16);
}
}
@@ -767,16 +888,21 @@ void edma_set_transfer_params(unsigned slot,
u16 acnt, u16 bcnt, u16 ccnt,
u16 bcnt_rld, enum sync_dimension sync_mode)
{
- if (slot < num_slots) {
- edma_parm_modify(PARM_LINK_BCNTRLD, slot,
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot < edma_info[ctlr]->num_slots) {
+ edma_parm_modify(ctlr, PARM_LINK_BCNTRLD, slot,
0x0000ffff, bcnt_rld << 16);
if (sync_mode == ASYNC)
- edma_parm_and(PARM_OPT, slot, ~SYNCDIM);
+ edma_parm_and(ctlr, PARM_OPT, slot, ~SYNCDIM);
else
- edma_parm_or(PARM_OPT, slot, SYNCDIM);
+ edma_parm_or(ctlr, PARM_OPT, slot, SYNCDIM);
/* Set the acount, bcount, ccount registers */
- edma_parm_write(PARM_A_B_CNT, slot, (bcnt << 16) | acnt);
- edma_parm_write(PARM_CCNT, slot, ccnt);
+ edma_parm_write(ctlr, PARM_A_B_CNT, slot, (bcnt << 16) | acnt);
+ edma_parm_write(ctlr, PARM_CCNT, slot, ccnt);
}
}
EXPORT_SYMBOL(edma_set_transfer_params);
@@ -790,11 +916,19 @@ EXPORT_SYMBOL(edma_set_transfer_params);
*/
void edma_link(unsigned from, unsigned to)
{
- if (from >= num_slots)
+ unsigned ctlr_from, ctlr_to;
+
+ ctlr_from = EDMA_CTLR(from);
+ from = EDMA_CHAN_SLOT(from);
+ ctlr_to = EDMA_CTLR(to);
+ to = EDMA_CHAN_SLOT(to);
+
+ if (from >= edma_info[ctlr_from]->num_slots)
return;
- if (to >= num_slots)
+ if (to >= edma_info[ctlr_to]->num_slots)
return;
- edma_parm_modify(PARM_LINK_BCNTRLD, from, 0xffff0000, PARM_OFFSET(to));
+ edma_parm_modify(ctlr_from, PARM_LINK_BCNTRLD, from, 0xffff0000,
+ PARM_OFFSET(to));
}
EXPORT_SYMBOL(edma_link);
@@ -807,9 +941,14 @@ EXPORT_SYMBOL(edma_link);
*/
void edma_unlink(unsigned from)
{
- if (from >= num_slots)
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(from);
+ from = EDMA_CHAN_SLOT(from);
+
+ if (from >= edma_info[ctlr]->num_slots)
return;
- edma_parm_or(PARM_LINK_BCNTRLD, from, 0xffff);
+ edma_parm_or(ctlr, PARM_LINK_BCNTRLD, from, 0xffff);
}
EXPORT_SYMBOL(edma_unlink);
@@ -829,9 +968,15 @@ EXPORT_SYMBOL(edma_unlink);
*/
void edma_write_slot(unsigned slot, const struct edmacc_param *param)
{
- if (slot >= num_slots)
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot >= edma_info[ctlr]->num_slots)
return;
- memcpy_toio(edmacc_regs_base + PARM_OFFSET(slot), param, PARM_SIZE);
+ memcpy_toio(edmacc_regs_base[ctlr] + PARM_OFFSET(slot), param,
+ PARM_SIZE);
}
EXPORT_SYMBOL(edma_write_slot);
@@ -845,9 +990,15 @@ EXPORT_SYMBOL(edma_write_slot);
*/
void edma_read_slot(unsigned slot, struct edmacc_param *param)
{
- if (slot >= num_slots)
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(slot);
+ slot = EDMA_CHAN_SLOT(slot);
+
+ if (slot >= edma_info[ctlr]->num_slots)
return;
- memcpy_fromio(param, edmacc_regs_base + PARM_OFFSET(slot), PARM_SIZE);
+ memcpy_fromio(param, edmacc_regs_base[ctlr] + PARM_OFFSET(slot),
+ PARM_SIZE);
}
EXPORT_SYMBOL(edma_read_slot);
@@ -864,10 +1015,15 @@ EXPORT_SYMBOL(edma_read_slot);
*/
void edma_pause(unsigned channel)
{
- if (channel < num_channels) {
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel < edma_info[ctlr]->num_channels) {
unsigned int mask = (1 << (channel & 0x1f));
- edma_shadow0_write_array(SH_EECR, channel >> 5, mask);
+ edma_shadow0_write_array(ctlr, SH_EECR, channel >> 5, mask);
}
}
EXPORT_SYMBOL(edma_pause);
@@ -880,10 +1036,15 @@ EXPORT_SYMBOL(edma_pause);
*/
void edma_resume(unsigned channel)
{
- if (channel < num_channels) {
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel < edma_info[ctlr]->num_channels) {
unsigned int mask = (1 << (channel & 0x1f));
- edma_shadow0_write_array(SH_EESR, channel >> 5, mask);
+ edma_shadow0_write_array(ctlr, SH_EESR, channel >> 5, mask);
}
}
EXPORT_SYMBOL(edma_resume);
@@ -901,28 +1062,33 @@ EXPORT_SYMBOL(edma_resume);
*/
int edma_start(unsigned channel)
{
- if (channel < num_channels) {
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel < edma_info[ctlr]->num_channels) {
int j = channel >> 5;
unsigned int mask = (1 << (channel & 0x1f));
/* EDMA channels without event association */
- if (test_bit(channel, edma_noevent)) {
+ if (test_bit(channel, edma_info[ctlr]->edma_noevent)) {
pr_debug("EDMA: ESR%d %08x\n", j,
- edma_shadow0_read_array(SH_ESR, j));
- edma_shadow0_write_array(SH_ESR, j, mask);
+ edma_shadow0_read_array(ctlr, SH_ESR, j));
+ edma_shadow0_write_array(ctlr, SH_ESR, j, mask);
return 0;
}
/* EDMA channel with event association */
pr_debug("EDMA: ER%d %08x\n", j,
- edma_shadow0_read_array(SH_ER, j));
+ edma_shadow0_read_array(ctlr, SH_ER, j));
/* Clear any pending error */
- edma_write_array(EDMA_EMCR, j, mask);
+ edma_write_array(ctlr, EDMA_EMCR, j, mask);
/* Clear any SER */
- edma_shadow0_write_array(SH_SECR, j, mask);
- edma_shadow0_write_array(SH_EESR, j, mask);
+ edma_shadow0_write_array(ctlr, SH_SECR, j, mask);
+ edma_shadow0_write_array(ctlr, SH_EESR, j, mask);
pr_debug("EDMA: EER%d %08x\n", j,
- edma_shadow0_read_array(SH_EER, j));
+ edma_shadow0_read_array(ctlr, SH_EER, j));
return 0;
}
@@ -941,17 +1107,22 @@ EXPORT_SYMBOL(edma_start);
*/
void edma_stop(unsigned channel)
{
- if (channel < num_channels) {
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel < edma_info[ctlr]->num_channels) {
int j = channel >> 5;
unsigned int mask = (1 << (channel & 0x1f));
- edma_shadow0_write_array(SH_EECR, j, mask);
- edma_shadow0_write_array(SH_ECR, j, mask);
- edma_shadow0_write_array(SH_SECR, j, mask);
- edma_write_array(EDMA_EMCR, j, mask);
+ edma_shadow0_write_array(ctlr, SH_EECR, j, mask);
+ edma_shadow0_write_array(ctlr, SH_ECR, j, mask);
+ edma_shadow0_write_array(ctlr, SH_SECR, j, mask);
+ edma_write_array(ctlr, EDMA_EMCR, j, mask);
pr_debug("EDMA: EER%d %08x\n", j,
- edma_shadow0_read_array(SH_EER, j));
+ edma_shadow0_read_array(ctlr, SH_EER, j));
/* REVISIT: consider guarding against inappropriate event
* chaining by overwriting with dummy_paramset.
@@ -975,18 +1146,23 @@ EXPORT_SYMBOL(edma_stop);
void edma_clean_channel(unsigned channel)
{
- if (channel < num_channels) {
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel < edma_info[ctlr]->num_channels) {
int j = (channel >> 5);
unsigned int mask = 1 << (channel & 0x1f);
pr_debug("EDMA: EMR%d %08x\n", j,
- edma_read_array(EDMA_EMR, j));
- edma_shadow0_write_array(SH_ECR, j, mask);
+ edma_read_array(ctlr, EDMA_EMR, j));
+ edma_shadow0_write_array(ctlr, SH_ECR, j, mask);
/* Clear the corresponding EMR bits */
- edma_write_array(EDMA_EMCR, j, mask);
+ edma_write_array(ctlr, EDMA_EMCR, j, mask);
/* Clear any SER */
- edma_shadow0_write_array(SH_SECR, j, mask);
- edma_write(EDMA_CCERRCLR, (1 << 16) | 0x3);
+ edma_shadow0_write_array(ctlr, SH_SECR, j, mask);
+ edma_write(ctlr, EDMA_CCERRCLR, (1 << 16) | 0x3);
}
}
EXPORT_SYMBOL(edma_clean_channel);
@@ -998,12 +1174,17 @@ EXPORT_SYMBOL(edma_clean_channel);
*/
void edma_clear_event(unsigned channel)
{
- if (channel >= num_channels)
+ unsigned ctlr;
+
+ ctlr = EDMA_CTLR(channel);
+ channel = EDMA_CHAN_SLOT(channel);
+
+ if (channel >= edma_info[ctlr]->num_channels)
return;
if (channel < 32)
- edma_write(EDMA_ECR, 1 << channel);
+ edma_write(ctlr, EDMA_ECR, 1 << channel);
else
- edma_write(EDMA_ECRH, 1 << (channel - 32));
+ edma_write(ctlr, EDMA_ECRH, 1 << (channel - 32));
}
EXPORT_SYMBOL(edma_clear_event);
@@ -1012,62 +1193,129 @@ EXPORT_SYMBOL(edma_clear_event);
static int __init edma_probe(struct platform_device *pdev)
{
struct edma_soc_info *info = pdev->dev.platform_data;
- int i;
- int status;
+ const s8 (*queue_priority_mapping)[2];
+ const s8 (*queue_tc_mapping)[2];
+ int i, j, found = 0;
+ int status = -1;
const s8 *noevent;
- int irq = 0, err_irq = 0;
- struct resource *r;
- resource_size_t len;
+ int irq[EDMA_MAX_CC] = {0, 0};
+ int err_irq[EDMA_MAX_CC] = {0, 0};
+ struct resource *r[EDMA_MAX_CC] = {NULL};
+ resource_size_t len[EDMA_MAX_CC];
+ char res_name[10];
+ char irq_name[10];
if (!info)
return -ENODEV;
- r = platform_get_resource_byname(pdev, IORESOURCE_MEM, "edma_cc");
- if (!r)
- return -ENODEV;
+ for (j = 0; j < EDMA_MAX_CC; j++) {
+ sprintf(res_name, "edma_cc%d", j);
+ r[j] = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+ res_name);
+ if (!r[j]) {
+ if (found)
+ break;
+ else
+ return -ENODEV;
+ } else
+ found = 1;
+
+ len[j] = resource_size(r[j]);
+
+ r[j] = request_mem_region(r[j]->start, len[j],
+ dev_name(&pdev->dev));
+ if (!r[j]) {
+ status = -EBUSY;
+ goto fail1;
+ }
- len = r->end - r->start + 1;
+ edmacc_regs_base[j] = ioremap(r[j]->start, len[j]);
+ if (!edmacc_regs_base[j]) {
+ status = -EBUSY;
+ goto fail1;
+ }
- r = request_mem_region(r->start, len, r->name);
- if (!r)
- return -EBUSY;
+ edma_info[j] = kmalloc(sizeof(struct edma), GFP_KERNEL);
+ if (!edma_info[j]) {
+ status = -ENOMEM;
+ goto fail1;
+ }
+ memset(edma_info[j], 0, sizeof(struct edma));
+
+ edma_info[j]->num_channels = min_t(unsigned, info[j].n_channel,
+ EDMA_MAX_DMACH);
+ edma_info[j]->num_slots = min_t(unsigned, info[j].n_slot,
+ EDMA_MAX_PARAMENTRY);
+ edma_info[j]->num_cc = min_t(unsigned, info[j].n_cc,
+ EDMA_MAX_CC);
+
+ dev_dbg(&pdev->dev, "DMA REG BASE ADDR=%p\n",
+ edmacc_regs_base[j]);
+
+ for (i = 0; i < edma_info[j]->num_slots; i++)
+ memcpy_toio(edmacc_regs_base[j] + PARM_OFFSET(i),
+ &dummy_paramset, PARM_SIZE);
+
+ noevent = info[j].noevent;
+ if (noevent) {
+ while (*noevent != -1)
+ set_bit(*noevent++, edma_info[j]->edma_noevent);
+ }
- edmacc_regs_base = ioremap(r->start, len);
- if (!edmacc_regs_base) {
- status = -EBUSY;
- goto fail1;
- }
+ sprintf(irq_name, "edma%d", j);
+ irq[j] = platform_get_irq_byname(pdev, irq_name);
+ edma_info[j]->irq_res_start = irq[j];
+ status = request_irq(irq[j], dma_irq_handler, 0, "edma",
+ &pdev->dev);
+ if (status < 0) {
+ dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
+ irq[j], status);
+ goto fail;
+ }
- num_channels = min_t(unsigned, info->n_channel, EDMA_MAX_DMACH);
- num_slots = min_t(unsigned, info->n_slot, EDMA_MAX_PARAMENTRY);
+ sprintf(irq_name, "edma%d_err", j);
+ err_irq[j] = platform_get_irq_byname(pdev, irq_name);
+ edma_info[j]->irq_res_end = err_irq[j];
+ status = request_irq(err_irq[j], dma_ccerr_handler, 0,
+ "edma_error", &pdev->dev);
+ if (status < 0) {
+ dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
+ err_irq[j], status);
+ goto fail;
+ }
- dev_dbg(&pdev->dev, "DMA REG BASE ADDR=%p\n", edmacc_regs_base);
+ /* Everything lives on transfer controller 1 until otherwise
+ * specified. This way, long transfers on the low priority queue
+ * started by the codec engine will not cause audio defects.
+ */
+ for (i = 0; i < edma_info[j]->num_channels; i++)
+ map_dmach_queue(j, i, EVENTQ_1);
- for (i = 0; i < num_slots; i++)
- memcpy_toio(edmacc_regs_base + PARM_OFFSET(i),
- &dummy_paramset, PARM_SIZE);
+ queue_tc_mapping = info[j].queue_tc_mapping;
+ queue_priority_mapping = info[j].queue_priority_mapping;
- noevent = info->noevent;
- if (noevent) {
- while (*noevent != -1)
- set_bit(*noevent++, edma_noevent);
- }
+ /* Event queue to TC mapping */
+ for (i = 0; queue_tc_mapping[i][0] != -1; i++)
+ map_queue_tc(j, queue_tc_mapping[i][0],
+ queue_tc_mapping[i][1]);
- irq = platform_get_irq(pdev, 0);
- status = request_irq(irq, dma_irq_handler, 0, "edma", &pdev->dev);
- if (status < 0) {
- dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
- irq, status);
- goto fail;
- }
+ /* Event queue priority mapping */
+ for (i = 0; queue_priority_mapping[i][0] != -1; i++)
+ assign_priority_to_queue(j,
+ queue_priority_mapping[i][0],
+ queue_priority_mapping[i][1]);
+
+ /* Map the channel to param entry if channel mapping logic
+ * exist
+ */
+ if (edma_read(j, EDMA_CCCFG) & CHMAP_EXIST)
+ map_dmach_param(j);
- err_irq = platform_get_irq(pdev, 1);
- status = request_irq(err_irq, dma_ccerr_handler, 0,
- "edma_error", &pdev->dev);
- if (status < 0) {
- dev_dbg(&pdev->dev, "request_irq %d failed --> %d\n",
- err_irq, status);
- goto fail;
+ for (i = 0; i < info[j].n_region; i++) {
+ edma_write_array2(j, EDMA_DRAE, i, 0, 0x0);
+ edma_write_array2(j, EDMA_DRAE, i, 1, 0x0);
+ edma_write_array(j, EDMA_QRAE, i, 0x0);
+ }
}
if (tc_errs_handled) {
@@ -1087,38 +1335,23 @@ static int __init edma_probe(struct platform_device *pdev)
}
}
- /* Everything lives on transfer controller 1 until otherwise specified.
- * This way, long transfers on the low priority queue
- * started by the codec engine will not cause audio defects.
- */
- for (i = 0; i < num_channels; i++)
- map_dmach_queue(i, EVENTQ_1);
-
- /* Event queue to TC mapping */
- for (i = 0; queue_tc_mapping[i][0] != -1; i++)
- map_queue_tc(queue_tc_mapping[i][0], queue_tc_mapping[i][1]);
-
- /* Event queue priority mapping */
- for (i = 0; queue_priority_mapping[i][0] != -1; i++)
- assign_priority_to_queue(queue_priority_mapping[i][0],
- queue_priority_mapping[i][1]);
-
- for (i = 0; i < info->n_region; i++) {
- edma_write_array2(EDMA_DRAE, i, 0, 0x0);
- edma_write_array2(EDMA_DRAE, i, 1, 0x0);
- edma_write_array(EDMA_QRAE, i, 0x0);
- }
-
return 0;
fail:
- if (err_irq)
- free_irq(err_irq, NULL);
- if (irq)
- free_irq(irq, NULL);
- iounmap(edmacc_regs_base);
+ for (i = 0; i < EDMA_MAX_CC; i++) {
+ if (err_irq[i])
+ free_irq(err_irq[i], &pdev->dev);
+ if (irq[i])
+ free_irq(irq[i], &pdev->dev);
+ }
fail1:
- release_mem_region(r->start, len);
+ for (i = 0; i < EDMA_MAX_CC; i++) {
+ if (r[i])
+ release_mem_region(r[i]->start, len[i]);
+ if (edmacc_regs_base[i])
+ iounmap(edmacc_regs_base[i]);
+ kfree(edma_info[i]);
+ }
return status;
}
diff --git a/arch/arm/mach-davinci/include/mach/edma.h b/arch/arm/mach-davinci/include/mach/edma.h
index 24a379239d7..ba2ebdd058a 100644
--- a/arch/arm/mach-davinci/include/mach/edma.h
+++ b/arch/arm/mach-davinci/include/mach/edma.h
@@ -170,6 +170,10 @@ enum sync_dimension {
ABSYNC = 1
};
+#define EDMA_CTLR_CHAN(ctlr, chan) (((ctlr) << 16) | (chan))
+#define EDMA_CTLR(i) ((i) >> 16)
+#define EDMA_CHAN_SLOT(i) ((i) & 0xffff)
+
#define EDMA_CHANNEL_ANY -1 /* for edma_alloc_channel() */
#define EDMA_SLOT_ANY -1 /* for edma_alloc_slot() */
@@ -180,7 +184,7 @@ int edma_alloc_channel(int channel,
void edma_free_channel(unsigned channel);
/* alloc/free parameter RAM slots */
-int edma_alloc_slot(int slot);
+int edma_alloc_slot(unsigned ctlr, int slot);
void edma_free_slot(unsigned slot);
/* calls that operate on part of a parameter RAM slot */
@@ -216,9 +220,12 @@ struct edma_soc_info {
unsigned n_region;
unsigned n_slot;
unsigned n_tc;
+ unsigned n_cc;
/* list of channels with no even trigger; terminated by "-1" */
const s8 *noevent;
+ const s8 (*queue_tc_mapping)[2];
+ const s8 (*queue_priority_mapping)[2];
};
#endif