提交 96583ddb 编写于 作者: D Daniel Vetter

Merge remote-tracking branch 'airlied/drm-next' into drm-intel-next-queued

Backmerge latest drm-next to pull in the s/fence/dma_fence/ rework,
needed before we merge more i915 fencing patches.
Signed-off-by: NDaniel Vetter <daniel.vetter@intel.com>
Silicon Image SiI8620 HDMI/MHL bridge bindings
Required properties:
- compatible: "sil,sii8620"
- reg: i2c address of the bridge
- cvcc10-supply: Digital Core Supply Voltage (1.0V)
- iovcc18-supply: I/O Supply Voltage (1.8V)
- interrupts, interrupt-parent: interrupt specifier of INT pin
- reset-gpios: gpio specifier of RESET pin
- clocks, clock-names: specification and name of "xtal" clock
- video interfaces: Device node can contain video interface port
node for HDMI encoder according to [1].
[1]: Documentation/devicetree/bindings/media/video-interfaces.txt
Example:
sii8620@39 {
reg = <0x39>;
compatible = "sil,sii8620";
cvcc10-supply = <&ldo36_reg>;
iovcc18-supply = <&ldo34_reg>;
interrupt-parent = <&gpf0>;
interrupts = <2 0>;
reset-gpio = <&gpv7 0 0>;
clocks = <&pmu_system_controller 0>;
clock-names = "xtal";
port {
mhl_to_hdmi: endpoint {
remote-endpoint = <&hdmi_to_mhl>;
};
};
};
......@@ -6,7 +6,7 @@
This document serves as a guide for device drivers writers on what the
sync_file API is, and how drivers can support it. Sync file is the carrier of
the fences(struct fence) that are needed to synchronize between drivers or
the fences(struct dma_fence) that are needed to synchronize between drivers or
across process boundaries.
The sync_file API is meant to be used to send and receive fence information
......@@ -32,9 +32,9 @@ in-fences and out-fences
Sync files can go either to or from userspace. When a sync_file is sent from
the driver to userspace we call the fences it contains 'out-fences'. They are
related to a buffer that the driver is processing or is going to process, so
the driver creates an out-fence to be able to notify, through fence_signal(),
when it has finished using (or processing) that buffer. Out-fences are fences
that the driver creates.
the driver creates an out-fence to be able to notify, through
dma_fence_signal(), when it has finished using (or processing) that buffer.
Out-fences are fences that the driver creates.
On the other hand if the driver receives fence(s) through a sync_file from
userspace we call these fence(s) 'in-fences'. Receiveing in-fences means that
......@@ -47,7 +47,7 @@ Creating Sync Files
When a driver needs to send an out-fence userspace it creates a sync_file.
Interface:
struct sync_file *sync_file_create(struct fence *fence);
struct sync_file *sync_file_create(struct dma_fence *fence);
The caller pass the out-fence and gets back the sync_file. That is just the
first step, next it needs to install an fd on sync_file->file. So it gets an
......@@ -72,11 +72,11 @@ of the Sync File to the kernel. The kernel can then retrieve the fences
from it.
Interface:
struct fence *sync_file_get_fence(int fd);
struct dma_fence *sync_file_get_fence(int fd);
The returned reference is owned by the caller and must be disposed of
afterwards using fence_put(). In case of error, a NULL is returned instead.
afterwards using dma_fence_put(). In case of error, a NULL is returned instead.
References:
[1] struct sync_file in include/linux/sync_file.h
......
......@@ -248,11 +248,11 @@ config DMA_SHARED_BUFFER
APIs extension; the file's descriptor can then be passed on to other
driver.
config FENCE_TRACE
bool "Enable verbose FENCE_TRACE messages"
config DMA_FENCE_TRACE
bool "Enable verbose DMA_FENCE_TRACE messages"
depends on DMA_SHARED_BUFFER
help
Enable the FENCE_TRACE printks. This will add extra
Enable the DMA_FENCE_TRACE printks. This will add extra
spam to the console log, but will make it easier to diagnose
lockup related problems for dma-buffers shared across multiple
devices.
......
......@@ -7,7 +7,7 @@ config SYNC_FILE
select DMA_SHARED_BUFFER
---help---
The Sync File Framework adds explicit syncronization via
userspace. It enables send/receive 'struct fence' objects to/from
userspace. It enables send/receive 'struct dma_fence' objects to/from
userspace via Sync File fds for synchronization between drivers via
userspace components. It has been ported from Android.
......
obj-y := dma-buf.o fence.o reservation.o seqno-fence.o fence-array.o
obj-y := dma-buf.o dma-fence.o dma-fence-array.o reservation.o seqno-fence.o
obj-$(CONFIG_SYNC_FILE) += sync_file.o
obj-$(CONFIG_SW_SYNC) += sw_sync.o sync_debug.o
......@@ -25,7 +25,7 @@
#include <linux/fs.h>
#include <linux/slab.h>
#include <linux/dma-buf.h>
#include <linux/fence.h>
#include <linux/dma-fence.h>
#include <linux/anon_inodes.h>
#include <linux/export.h>
#include <linux/debugfs.h>
......@@ -124,7 +124,7 @@ static loff_t dma_buf_llseek(struct file *file, loff_t offset, int whence)
return base + offset;
}
static void dma_buf_poll_cb(struct fence *fence, struct fence_cb *cb)
static void dma_buf_poll_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
{
struct dma_buf_poll_cb_t *dcb = (struct dma_buf_poll_cb_t *)cb;
unsigned long flags;
......@@ -140,7 +140,7 @@ static unsigned int dma_buf_poll(struct file *file, poll_table *poll)
struct dma_buf *dmabuf;
struct reservation_object *resv;
struct reservation_object_list *fobj;
struct fence *fence_excl;
struct dma_fence *fence_excl;
unsigned long events;
unsigned shared_count, seq;
......@@ -187,20 +187,20 @@ static unsigned int dma_buf_poll(struct file *file, poll_table *poll)
spin_unlock_irq(&dmabuf->poll.lock);
if (events & pevents) {
if (!fence_get_rcu(fence_excl)) {
if (!dma_fence_get_rcu(fence_excl)) {
/* force a recheck */
events &= ~pevents;
dma_buf_poll_cb(NULL, &dcb->cb);
} else if (!fence_add_callback(fence_excl, &dcb->cb,
dma_buf_poll_cb)) {
} else if (!dma_fence_add_callback(fence_excl, &dcb->cb,
dma_buf_poll_cb)) {
events &= ~pevents;
fence_put(fence_excl);
dma_fence_put(fence_excl);
} else {
/*
* No callback queued, wake up any additional
* waiters.
*/
fence_put(fence_excl);
dma_fence_put(fence_excl);
dma_buf_poll_cb(NULL, &dcb->cb);
}
}
......@@ -222,9 +222,9 @@ static unsigned int dma_buf_poll(struct file *file, poll_table *poll)
goto out;
for (i = 0; i < shared_count; ++i) {
struct fence *fence = rcu_dereference(fobj->shared[i]);
struct dma_fence *fence = rcu_dereference(fobj->shared[i]);
if (!fence_get_rcu(fence)) {
if (!dma_fence_get_rcu(fence)) {
/*
* fence refcount dropped to zero, this means
* that fobj has been freed
......@@ -235,13 +235,13 @@ static unsigned int dma_buf_poll(struct file *file, poll_table *poll)
dma_buf_poll_cb(NULL, &dcb->cb);
break;
}
if (!fence_add_callback(fence, &dcb->cb,
dma_buf_poll_cb)) {
fence_put(fence);
if (!dma_fence_add_callback(fence, &dcb->cb,
dma_buf_poll_cb)) {
dma_fence_put(fence);
events &= ~POLLOUT;
break;
}
fence_put(fence);
dma_fence_put(fence);
}
/* No callback queued, wake up any additional waiters. */
......
/*
* fence-array: aggregate fences to be waited together
* dma-fence-array: aggregate fences to be waited together
*
* Copyright (C) 2016 Collabora Ltd
* Copyright (C) 2016 Advanced Micro Devices, Inc.
......@@ -19,35 +19,34 @@
#include <linux/export.h>
#include <linux/slab.h>
#include <linux/fence-array.h>
#include <linux/dma-fence-array.h>
static void fence_array_cb_func(struct fence *f, struct fence_cb *cb);
static const char *fence_array_get_driver_name(struct fence *fence)
static const char *dma_fence_array_get_driver_name(struct dma_fence *fence)
{
return "fence_array";
return "dma_fence_array";
}
static const char *fence_array_get_timeline_name(struct fence *fence)
static const char *dma_fence_array_get_timeline_name(struct dma_fence *fence)
{
return "unbound";
}
static void fence_array_cb_func(struct fence *f, struct fence_cb *cb)
static void dma_fence_array_cb_func(struct dma_fence *f,
struct dma_fence_cb *cb)
{
struct fence_array_cb *array_cb =
container_of(cb, struct fence_array_cb, cb);
struct fence_array *array = array_cb->array;
struct dma_fence_array_cb *array_cb =
container_of(cb, struct dma_fence_array_cb, cb);
struct dma_fence_array *array = array_cb->array;
if (atomic_dec_and_test(&array->num_pending))
fence_signal(&array->base);
fence_put(&array->base);
dma_fence_signal(&array->base);
dma_fence_put(&array->base);
}
static bool fence_array_enable_signaling(struct fence *fence)
static bool dma_fence_array_enable_signaling(struct dma_fence *fence)
{
struct fence_array *array = to_fence_array(fence);
struct fence_array_cb *cb = (void *)(&array[1]);
struct dma_fence_array *array = to_dma_fence_array(fence);
struct dma_fence_array_cb *cb = (void *)(&array[1]);
unsigned i;
for (i = 0; i < array->num_fences; ++i) {
......@@ -60,10 +59,10 @@ static bool fence_array_enable_signaling(struct fence *fence)
* until we signal the array as complete (but that is now
* insufficient).
*/
fence_get(&array->base);
if (fence_add_callback(array->fences[i], &cb[i].cb,
fence_array_cb_func)) {
fence_put(&array->base);
dma_fence_get(&array->base);
if (dma_fence_add_callback(array->fences[i], &cb[i].cb,
dma_fence_array_cb_func)) {
dma_fence_put(&array->base);
if (atomic_dec_and_test(&array->num_pending))
return false;
}
......@@ -72,69 +71,71 @@ static bool fence_array_enable_signaling(struct fence *fence)
return true;
}
static bool fence_array_signaled(struct fence *fence)
static bool dma_fence_array_signaled(struct dma_fence *fence)
{
struct fence_array *array = to_fence_array(fence);
struct dma_fence_array *array = to_dma_fence_array(fence);
return atomic_read(&array->num_pending) <= 0;
}
static void fence_array_release(struct fence *fence)
static void dma_fence_array_release(struct dma_fence *fence)
{
struct fence_array *array = to_fence_array(fence);
struct dma_fence_array *array = to_dma_fence_array(fence);
unsigned i;
for (i = 0; i < array->num_fences; ++i)
fence_put(array->fences[i]);
dma_fence_put(array->fences[i]);
kfree(array->fences);
fence_free(fence);
dma_fence_free(fence);
}
const struct fence_ops fence_array_ops = {
.get_driver_name = fence_array_get_driver_name,
.get_timeline_name = fence_array_get_timeline_name,
.enable_signaling = fence_array_enable_signaling,
.signaled = fence_array_signaled,
.wait = fence_default_wait,
.release = fence_array_release,
const struct dma_fence_ops dma_fence_array_ops = {
.get_driver_name = dma_fence_array_get_driver_name,
.get_timeline_name = dma_fence_array_get_timeline_name,
.enable_signaling = dma_fence_array_enable_signaling,
.signaled = dma_fence_array_signaled,
.wait = dma_fence_default_wait,
.release = dma_fence_array_release,
};
EXPORT_SYMBOL(fence_array_ops);
EXPORT_SYMBOL(dma_fence_array_ops);
/**
* fence_array_create - Create a custom fence array
* dma_fence_array_create - Create a custom fence array
* @num_fences: [in] number of fences to add in the array
* @fences: [in] array containing the fences
* @context: [in] fence context to use
* @seqno: [in] sequence number to use
* @signal_on_any: [in] signal on any fence in the array
*
* Allocate a fence_array object and initialize the base fence with fence_init().
* Allocate a dma_fence_array object and initialize the base fence with
* dma_fence_init().
* In case of error it returns NULL.
*
* The caller should allocate the fences array with num_fences size
* and fill it with the fences it wants to add to the object. Ownership of this
* array is taken and fence_put() is used on each fence on release.
* array is taken and dma_fence_put() is used on each fence on release.
*
* If @signal_on_any is true the fence array signals if any fence in the array
* signals, otherwise it signals when all fences in the array signal.
*/
struct fence_array *fence_array_create(int num_fences, struct fence **fences,
u64 context, unsigned seqno,
bool signal_on_any)
struct dma_fence_array *dma_fence_array_create(int num_fences,
struct dma_fence **fences,
u64 context, unsigned seqno,
bool signal_on_any)
{
struct fence_array *array;
struct dma_fence_array *array;
size_t size = sizeof(*array);
/* Allocate the callback structures behind the array. */
size += num_fences * sizeof(struct fence_array_cb);
size += num_fences * sizeof(struct dma_fence_array_cb);
array = kzalloc(size, GFP_KERNEL);
if (!array)
return NULL;
spin_lock_init(&array->lock);
fence_init(&array->base, &fence_array_ops, &array->lock,
context, seqno);
dma_fence_init(&array->base, &dma_fence_array_ops, &array->lock,
context, seqno);
array->num_fences = num_fences;
atomic_set(&array->num_pending, signal_on_any ? 1 : num_fences);
......@@ -142,4 +143,4 @@ struct fence_array *fence_array_create(int num_fences, struct fence **fences,
return array;
}
EXPORT_SYMBOL(fence_array_create);
EXPORT_SYMBOL(dma_fence_array_create);
......@@ -21,13 +21,13 @@
#include <linux/slab.h>
#include <linux/export.h>
#include <linux/atomic.h>
#include <linux/fence.h>
#include <linux/dma-fence.h>
#define CREATE_TRACE_POINTS
#include <trace/events/fence.h>
#include <trace/events/dma_fence.h>
EXPORT_TRACEPOINT_SYMBOL(fence_annotate_wait_on);
EXPORT_TRACEPOINT_SYMBOL(fence_emit);
EXPORT_TRACEPOINT_SYMBOL(dma_fence_annotate_wait_on);
EXPORT_TRACEPOINT_SYMBOL(dma_fence_emit);
/*
* fence context counter: each execution context should have its own
......@@ -35,39 +35,41 @@ EXPORT_TRACEPOINT_SYMBOL(fence_emit);
* context or not. One device can have multiple separate contexts,
* and they're used if some engine can run independently of another.
*/
static atomic64_t fence_context_counter = ATOMIC64_INIT(0);
static atomic64_t dma_fence_context_counter = ATOMIC64_INIT(0);
/**
* fence_context_alloc - allocate an array of fence contexts
* dma_fence_context_alloc - allocate an array of fence contexts
* @num: [in] amount of contexts to allocate
*
* This function will return the first index of the number of fences allocated.
* The fence context is used for setting fence->context to a unique number.
*/
u64 fence_context_alloc(unsigned num)
u64 dma_fence_context_alloc(unsigned num)
{
BUG_ON(!num);
return atomic64_add_return(num, &fence_context_counter) - num;
return atomic64_add_return(num, &dma_fence_context_counter) - num;
}
EXPORT_SYMBOL(fence_context_alloc);
EXPORT_SYMBOL(dma_fence_context_alloc);
/**
* fence_signal_locked - signal completion of a fence
* dma_fence_signal_locked - signal completion of a fence
* @fence: the fence to signal
*
* Signal completion for software callbacks on a fence, this will unblock
* fence_wait() calls and run all the callbacks added with
* fence_add_callback(). Can be called multiple times, but since a fence
* dma_fence_wait() calls and run all the callbacks added with
* dma_fence_add_callback(). Can be called multiple times, but since a fence
* can only go from unsignaled to signaled state, it will only be effective
* the first time.
*
* Unlike fence_signal, this function must be called with fence->lock held.
* Unlike dma_fence_signal, this function must be called with fence->lock held.
*/
int fence_signal_locked(struct fence *fence)
int dma_fence_signal_locked(struct dma_fence *fence)
{
struct fence_cb *cur, *tmp;
struct dma_fence_cb *cur, *tmp;
int ret = 0;
lockdep_assert_held(fence->lock);
if (WARN_ON(!fence))
return -EINVAL;
......@@ -76,15 +78,15 @@ int fence_signal_locked(struct fence *fence)
smp_mb__before_atomic();
}
if (test_and_set_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
if (test_and_set_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
ret = -EINVAL;
/*
* we might have raced with the unlocked fence_signal,
* we might have raced with the unlocked dma_fence_signal,
* still run through all callbacks
*/
} else
trace_fence_signaled(fence);
trace_dma_fence_signaled(fence);
list_for_each_entry_safe(cur, tmp, &fence->cb_list, node) {
list_del_init(&cur->node);
......@@ -92,19 +94,19 @@ int fence_signal_locked(struct fence *fence)
}
return ret;
}
EXPORT_SYMBOL(fence_signal_locked);
EXPORT_SYMBOL(dma_fence_signal_locked);
/**
* fence_signal - signal completion of a fence
* dma_fence_signal - signal completion of a fence
* @fence: the fence to signal
*
* Signal completion for software callbacks on a fence, this will unblock
* fence_wait() calls and run all the callbacks added with
* fence_add_callback(). Can be called multiple times, but since a fence
* dma_fence_wait() calls and run all the callbacks added with
* dma_fence_add_callback(). Can be called multiple times, but since a fence
* can only go from unsignaled to signaled state, it will only be effective
* the first time.
*/
int fence_signal(struct fence *fence)
int dma_fence_signal(struct dma_fence *fence)
{
unsigned long flags;
......@@ -116,13 +118,13 @@ int fence_signal(struct fence *fence)
smp_mb__before_atomic();
}
if (test_and_set_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags))
if (test_and_set_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
return -EINVAL;
trace_fence_signaled(fence);
trace_dma_fence_signaled(fence);
if (test_bit(FENCE_FLAG_ENABLE_SIGNAL_BIT, &fence->flags)) {
struct fence_cb *cur, *tmp;
if (test_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT, &fence->flags)) {
struct dma_fence_cb *cur, *tmp;
spin_lock_irqsave(fence->lock, flags);
list_for_each_entry_safe(cur, tmp, &fence->cb_list, node) {
......@@ -133,10 +135,10 @@ int fence_signal(struct fence *fence)
}
return 0;
}
EXPORT_SYMBOL(fence_signal);
EXPORT_SYMBOL(dma_fence_signal);
/**
* fence_wait_timeout - sleep until the fence gets signaled
* dma_fence_wait_timeout - sleep until the fence gets signaled
* or until timeout elapses
* @fence: [in] the fence to wait on
* @intr: [in] if true, do an interruptible wait
......@@ -152,7 +154,7 @@ EXPORT_SYMBOL(fence_signal);
* freed before return, resulting in undefined behavior.
*/
signed long
fence_wait_timeout(struct fence *fence, bool intr, signed long timeout)
dma_fence_wait_timeout(struct dma_fence *fence, bool intr, signed long timeout)
{
signed long ret;
......@@ -160,70 +162,71 @@ fence_wait_timeout(struct fence *fence, bool intr, signed long timeout)
return -EINVAL;
if (timeout == 0)
return fence_is_signaled(fence);
return dma_fence_is_signaled(fence);
trace_fence_wait_start(fence);
trace_dma_fence_wait_start(fence);
ret = fence->ops->wait(fence, intr, timeout);
trace_fence_wait_end(fence);
trace_dma_fence_wait_end(fence);
return ret;
}
EXPORT_SYMBOL(fence_wait_timeout);
EXPORT_SYMBOL(dma_fence_wait_timeout);
void fence_release(struct kref *kref)
void dma_fence_release(struct kref *kref)
{
struct fence *fence =
container_of(kref, struct fence, refcount);
struct dma_fence *fence =
container_of(kref, struct dma_fence, refcount);
trace_fence_destroy(fence);
trace_dma_fence_destroy(fence);
BUG_ON(!list_empty(&fence->cb_list));
if (fence->ops->release)
fence->ops->release(fence);
else
fence_free(fence);
dma_fence_free(fence);
}
EXPORT_SYMBOL(fence_release);
EXPORT_SYMBOL(dma_fence_release);
void fence_free(struct fence *fence)
void dma_fence_free(struct dma_fence *fence)
{
kfree_rcu(fence, rcu);
}
EXPORT_SYMBOL(fence_free);
EXPORT_SYMBOL(dma_fence_free);
/**
* fence_enable_sw_signaling - enable signaling on fence
* dma_fence_enable_sw_signaling - enable signaling on fence
* @fence: [in] the fence to enable
*
* this will request for sw signaling to be enabled, to make the fence
* complete as soon as possible
*/
void fence_enable_sw_signaling(struct fence *fence)
void dma_fence_enable_sw_signaling(struct dma_fence *fence)
{
unsigned long flags;
if (!test_and_set_bit(FENCE_FLAG_ENABLE_SIGNAL_BIT, &fence->flags) &&
!test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
trace_fence_enable_signal(fence);
if (!test_and_set_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT,
&fence->flags) &&
!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
trace_dma_fence_enable_signal(fence);
spin_lock_irqsave(fence->lock, flags);
if (!fence->ops->enable_signaling(fence))
fence_signal_locked(fence);
dma_fence_signal_locked(fence);
spin_unlock_irqrestore(fence->lock, flags);
}
}
EXPORT_SYMBOL(fence_enable_sw_signaling);
EXPORT_SYMBOL(dma_fence_enable_sw_signaling);
/**
* fence_add_callback - add a callback to be called when the fence
* dma_fence_add_callback - add a callback to be called when the fence
* is signaled
* @fence: [in] the fence to wait on
* @cb: [in] the callback to register
* @func: [in] the function to call
*
* cb will be initialized by fence_add_callback, no initialization
* cb will be initialized by dma_fence_add_callback, no initialization
* by the caller is required. Any number of callbacks can be registered
* to a fence, but a callback can only be registered to one fence at a time.
*
......@@ -232,15 +235,15 @@ EXPORT_SYMBOL(fence_enable_sw_signaling);
* *not* call the callback)
*
* Add a software callback to the fence. Same restrictions apply to
* refcount as it does to fence_wait, however the caller doesn't need to
* refcount as it does to dma_fence_wait, however the caller doesn't need to
* keep a refcount to fence afterwards: when software access is enabled,
* the creator of the fence is required to keep the fence alive until
* after it signals with fence_signal. The callback itself can be called
* after it signals with dma_fence_signal. The callback itself can be called
* from irq context.
*
*/
int fence_add_callback(struct fence *fence, struct fence_cb *cb,
fence_func_t func)
int dma_fence_add_callback(struct dma_fence *fence, struct dma_fence_cb *cb,
dma_fence_func_t func)
{
unsigned long flags;
int ret = 0;
......@@ -249,22 +252,23 @@ int fence_add_callback(struct fence *fence, struct fence_cb *cb,
if (WARN_ON(!fence || !func))
return -EINVAL;
if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags)) {
INIT_LIST_HEAD(&cb->node);
return -ENOENT;
}
spin_lock_irqsave(fence->lock, flags);
was_set = test_and_set_bit(FENCE_FLAG_ENABLE_SIGNAL_BIT, &fence->flags);
was_set = test_and_set_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT,
&fence->flags);
if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags))
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
ret = -ENOENT;
else if (!was_set) {
trace_fence_enable_signal(fence);
trace_dma_fence_enable_signal(fence);
if (!fence->ops->enable_signaling(fence)) {
fence_signal_locked(fence);
dma_fence_signal_locked(fence);
ret = -ENOENT;
}
}
......@@ -278,10 +282,10 @@ int fence_add_callback(struct fence *fence, struct fence_cb *cb,
return ret;
}
EXPORT_SYMBOL(fence_add_callback);
EXPORT_SYMBOL(dma_fence_add_callback);
/**
* fence_remove_callback - remove a callback from the signaling list
* dma_fence_remove_callback - remove a callback from the signaling list
* @fence: [in] the fence to wait on
* @cb: [in] the callback to remove
*
......@@ -296,7 +300,7 @@ EXPORT_SYMBOL(fence_add_callback);
* with a reference held to the fence.
*/
bool
fence_remove_callback(struct fence *fence, struct fence_cb *cb)
dma_fence_remove_callback(struct dma_fence *fence, struct dma_fence_cb *cb)
{
unsigned long flags;
bool ret;
......@@ -311,15 +315,15 @@ fence_remove_callback(struct fence *fence, struct fence_cb *cb)
return ret;
}
EXPORT_SYMBOL(fence_remove_callback);
EXPORT_SYMBOL(dma_fence_remove_callback);
struct default_wait_cb {
struct fence_cb base;
struct dma_fence_cb base;
struct task_struct *task;
};
static void
fence_default_wait_cb(struct fence *fence, struct fence_cb *cb)
dma_fence_default_wait_cb(struct dma_fence *fence, struct dma_fence_cb *cb)
{
struct default_wait_cb *wait =
container_of(cb, struct default_wait_cb, base);
......@@ -328,7 +332,7 @@ fence_default_wait_cb(struct fence *fence, struct fence_cb *cb)
}
/**
* fence_default_wait - default sleep until the fence gets signaled
* dma_fence_default_wait - default sleep until the fence gets signaled
* or until timeout elapses
* @fence: [in] the fence to wait on
* @intr: [in] if true, do an interruptible wait
......@@ -338,14 +342,14 @@ fence_default_wait_cb(struct fence *fence, struct fence_cb *cb)
* remaining timeout in jiffies on success.
*/
signed long
fence_default_wait(struct fence *fence, bool intr, signed long timeout)
dma_fence_default_wait(struct dma_fence *fence, bool intr, signed long timeout)
{
struct default_wait_cb cb;
unsigned long flags;
signed long ret = timeout;
bool was_set;
if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags))
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
return timeout;
spin_lock_irqsave(fence->lock, flags);
......@@ -355,25 +359,26 @@ fence_default_wait(struct fence *fence, bool intr, signed long timeout)
goto out;
}
was_set = test_and_set_bit(FENCE_FLAG_ENABLE_SIGNAL_BIT, &fence->flags);
was_set = test_and_set_bit(DMA_FENCE_FLAG_ENABLE_SIGNAL_BIT,
&fence->flags);
if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags))
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
goto out;
if (!was_set) {
trace_fence_enable_signal(fence);
trace_dma_fence_enable_signal(fence);
if (!fence->ops->enable_signaling(fence)) {
fence_signal_locked(fence);
dma_fence_signal_locked(fence);
goto out;
}
}
cb.base.func = fence_default_wait_cb;
cb.base.func = dma_fence_default_wait_cb;
cb.task = current;
list_add(&cb.base.node, &fence->cb_list);
while (!test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags) && ret > 0) {
while (!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags) && ret > 0) {
if (intr)
__set_current_state(TASK_INTERRUPTIBLE);
else
......@@ -395,23 +400,23 @@ fence_default_wait(struct fence *fence, bool intr, signed long timeout)
spin_unlock_irqrestore(fence->lock, flags);
return ret;
}
EXPORT_SYMBOL(fence_default_wait);
EXPORT_SYMBOL(dma_fence_default_wait);
static bool
fence_test_signaled_any(struct fence **fences, uint32_t count)
dma_fence_test_signaled_any(struct dma_fence **fences, uint32_t count)
{
int i;
for (i = 0; i < count; ++i) {
struct fence *fence = fences[i];
if (test_bit(FENCE_FLAG_SIGNALED_BIT, &fence->flags))
struct dma_fence *fence = fences[i];
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &fence->flags))
return true;
}
return false;
}
/**
* fence_wait_any_timeout - sleep until any fence gets signaled
* dma_fence_wait_any_timeout - sleep until any fence gets signaled
* or until timeout elapses
* @fences: [in] array of fences to wait on
* @count: [in] number of fences to wait on
......@@ -427,8 +432,8 @@ fence_test_signaled_any(struct fence **fences, uint32_t count)
* fence might be freed before return, resulting in undefined behavior.
*/
signed long
fence_wait_any_timeout(struct fence **fences, uint32_t count,
bool intr, signed long timeout)
dma_fence_wait_any_timeout(struct dma_fence **fences, uint32_t count,
bool intr, signed long timeout)
{
struct default_wait_cb *cb;
signed long ret = timeout;
......@@ -439,7 +444,7 @@ fence_wait_any_timeout(struct fence **fences, uint32_t count,
if (timeout == 0) {
for (i = 0; i < count; ++i)
if (fence_is_signaled(fences[i]))
if (dma_fence_is_signaled(fences[i]))
return 1;
return 0;
......@@ -452,16 +457,16 @@ fence_wait_any_timeout(struct fence **fences, uint32_t count,
}
for (i = 0; i < count; ++i) {
struct fence *fence = fences[i];
struct dma_fence *fence = fences[i];
if (fence->ops->wait != fence_default_wait) {
if (fence->ops->wait != dma_fence_default_wait) {
ret = -EINVAL;
goto fence_rm_cb;
}
cb[i].task = current;
if (fence_add_callback(fence, &cb[i].base,
fence_default_wait_cb)) {
if (dma_fence_add_callback(fence, &cb[i].base,
dma_fence_default_wait_cb)) {
/* This fence is already signaled */
goto fence_rm_cb;
}
......@@ -473,7 +478,7 @@ fence_wait_any_timeout(struct fence **fences, uint32_t count,
else
set_current_state(TASK_UNINTERRUPTIBLE);
if (fence_test_signaled_any(fences, count))
if (dma_fence_test_signaled_any(fences, count))
break;
ret = schedule_timeout(ret);
......@@ -486,34 +491,34 @@ fence_wait_any_timeout(struct fence **fences, uint32_t count,
fence_rm_cb:
while (i-- > 0)
fence_remove_callback(fences[i], &cb[i].base);
dma_fence_remove_callback(fences[i], &cb[i].base);
err_free_cb:
kfree(cb);
return ret;
}
EXPORT_SYMBOL(fence_wait_any_timeout);
EXPORT_SYMBOL(dma_fence_wait_any_timeout);
/**
* fence_init - Initialize a custom fence.
* dma_fence_init - Initialize a custom fence.
* @fence: [in] the fence to initialize
* @ops: [in] the fence_ops for operations on this fence
* @ops: [in] the dma_fence_ops for operations on this fence
* @lock: [in] the irqsafe spinlock to use for locking this fence
* @context: [in] the execution context this fence is run on
* @seqno: [in] a linear increasing sequence number for this context
*
* Initializes an allocated fence, the caller doesn't have to keep its
* refcount after committing with this fence, but it will need to hold a
* refcount again if fence_ops.enable_signaling gets called. This can
* refcount again if dma_fence_ops.enable_signaling gets called. This can
* be used for other implementing other types of fence.
*
* context and seqno are used for easy comparison between fences, allowing
* to check which fence is later by simply using fence_later.
* to check which fence is later by simply using dma_fence_later.
*/
void
fence_init(struct fence *fence, const struct fence_ops *ops,
spinlock_t *lock, u64 context, unsigned seqno)
dma_fence_init(struct dma_fence *fence, const struct dma_fence_ops *ops,
spinlock_t *lock, u64 context, unsigned seqno)
{
BUG_ON(!lock);
BUG_ON(!ops || !ops->wait || !ops->enable_signaling ||
......@@ -527,6 +532,6 @@ fence_init(struct fence *fence, const struct fence_ops *ops,
fence->seqno = seqno;
fence->flags = 0UL;
trace_fence_init(fence);
trace_dma_fence_init(fence);
}
EXPORT_SYMBOL(fence_init);
EXPORT_SYMBOL(dma_fence_init);
......@@ -102,17 +102,17 @@ EXPORT_SYMBOL(reservation_object_reserve_shared);
static void
reservation_object_add_shared_inplace(struct reservation_object *obj,
struct reservation_object_list *fobj,
struct fence *fence)
struct dma_fence *fence)
{
u32 i;
fence_get(fence);
dma_fence_get(fence);
preempt_disable();
write_seqcount_begin(&obj->seq);
for (i = 0; i < fobj->shared_count; ++i) {
struct fence *old_fence;
struct dma_fence *old_fence;
old_fence = rcu_dereference_protected(fobj->shared[i],
reservation_object_held(obj));
......@@ -123,7 +123,7 @@ reservation_object_add_shared_inplace(struct reservation_object *obj,
write_seqcount_end(&obj->seq);
preempt_enable();
fence_put(old_fence);
dma_fence_put(old_fence);
return;
}
}
......@@ -143,12 +143,12 @@ static void
reservation_object_add_shared_replace(struct reservation_object *obj,
struct reservation_object_list *old,
struct reservation_object_list *fobj,
struct fence *fence)
struct dma_fence *fence)
{
unsigned i;
struct fence *old_fence = NULL;
struct dma_fence *old_fence = NULL;
fence_get(fence);
dma_fence_get(fence);
if (!old) {
RCU_INIT_POINTER(fobj->shared[0], fence);
......@@ -165,7 +165,7 @@ reservation_object_add_shared_replace(struct reservation_object *obj,
fobj->shared_count = old->shared_count;
for (i = 0; i < old->shared_count; ++i) {
struct fence *check;
struct dma_fence *check;
check = rcu_dereference_protected(old->shared[i],
reservation_object_held(obj));
......@@ -196,7 +196,7 @@ reservation_object_add_shared_replace(struct reservation_object *obj,
kfree_rcu(old, rcu);
if (old_fence)
fence_put(old_fence);
dma_fence_put(old_fence);
}
/**
......@@ -208,7 +208,7 @@ reservation_object_add_shared_replace(struct reservation_object *obj,
* reservation_object_reserve_shared() has been called.
*/
void reservation_object_add_shared_fence(struct reservation_object *obj,
struct fence *fence)
struct dma_fence *fence)
{
struct reservation_object_list *old, *fobj = obj->staged;
......@@ -231,9 +231,9 @@ EXPORT_SYMBOL(reservation_object_add_shared_fence);
* Add a fence to the exclusive slot. The obj->lock must be held.
*/
void reservation_object_add_excl_fence(struct reservation_object *obj,
struct fence *fence)
struct dma_fence *fence)
{
struct fence *old_fence = reservation_object_get_excl(obj);
struct dma_fence *old_fence = reservation_object_get_excl(obj);
struct reservation_object_list *old;
u32 i = 0;
......@@ -242,7 +242,7 @@ void reservation_object_add_excl_fence(struct reservation_object *obj,
i = old->shared_count;
if (fence)
fence_get(fence);
dma_fence_get(fence);
preempt_disable();
write_seqcount_begin(&obj->seq);
......@@ -255,11 +255,11 @@ void reservation_object_add_excl_fence(struct reservation_object *obj,
/* inplace update, no shared fences */
while (i--)
fence_put(rcu_dereference_protected(old->shared[i],
dma_fence_put(rcu_dereference_protected(old->shared[i],
reservation_object_held(obj)));
if (old_fence)
fence_put(old_fence);
dma_fence_put(old_fence);
}
EXPORT_SYMBOL(reservation_object_add_excl_fence);
......@@ -276,12 +276,12 @@ EXPORT_SYMBOL(reservation_object_add_excl_fence);
* Zero or -errno
*/
int reservation_object_get_fences_rcu(struct reservation_object *obj,
struct fence **pfence_excl,
struct dma_fence **pfence_excl,
unsigned *pshared_count,
struct fence ***pshared)
struct dma_fence ***pshared)
{
struct fence **shared = NULL;
struct fence *fence_excl;
struct dma_fence **shared = NULL;
struct dma_fence *fence_excl;
unsigned int shared_count;
int ret = 1;
......@@ -296,12 +296,12 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj,
seq = read_seqcount_begin(&obj->seq);
fence_excl = rcu_dereference(obj->fence_excl);
if (fence_excl && !fence_get_rcu(fence_excl))
if (fence_excl && !dma_fence_get_rcu(fence_excl))
goto unlock;
fobj = rcu_dereference(obj->fence);
if (fobj) {
struct fence **nshared;
struct dma_fence **nshared;
size_t sz = sizeof(*shared) * fobj->shared_max;
nshared = krealloc(shared, sz,
......@@ -322,15 +322,15 @@ int reservation_object_get_fences_rcu(struct reservation_object *obj,
for (i = 0; i < shared_count; ++i) {
shared[i] = rcu_dereference(fobj->shared[i]);
if (!fence_get_rcu(shared[i]))
if (!dma_fence_get_rcu(shared[i]))
break;
}
}
if (i != shared_count || read_seqcount_retry(&obj->seq, seq)) {
while (i--)
fence_put(shared[i]);
fence_put(fence_excl);
dma_fence_put(shared[i]);
dma_fence_put(fence_excl);
goto unlock;
}
......@@ -368,7 +368,7 @@ long reservation_object_wait_timeout_rcu(struct reservation_object *obj,
bool wait_all, bool intr,
unsigned long timeout)
{
struct fence *fence;
struct dma_fence *fence;
unsigned seq, shared_count, i = 0;
long ret = timeout;
......@@ -389,16 +389,17 @@ long reservation_object_wait_timeout_rcu(struct reservation_object *obj,
shared_count = fobj->shared_count;
for (i = 0; i < shared_count; ++i) {
struct fence *lfence = rcu_dereference(fobj->shared[i]);
struct dma_fence *lfence = rcu_dereference(fobj->shared[i]);
if (test_bit(FENCE_FLAG_SIGNALED_BIT, &lfence->flags))
if (test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
&lfence->flags))
continue;
if (!fence_get_rcu(lfence))
if (!dma_fence_get_rcu(lfence))
goto unlock_retry;
if (fence_is_signaled(lfence)) {
fence_put(lfence);
if (dma_fence_is_signaled(lfence)) {
dma_fence_put(lfence);
continue;
}
......@@ -408,15 +409,16 @@ long reservation_object_wait_timeout_rcu(struct reservation_object *obj,
}
if (!shared_count) {
struct fence *fence_excl = rcu_dereference(obj->fence_excl);
struct dma_fence *fence_excl = rcu_dereference(obj->fence_excl);
if (fence_excl &&
!test_bit(FENCE_FLAG_SIGNALED_BIT, &fence_excl->flags)) {
if (!fence_get_rcu(fence_excl))
!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT,
&fence_excl->flags)) {
if (!dma_fence_get_rcu(fence_excl))
goto unlock_retry;
if (fence_is_signaled(fence_excl))
fence_put(fence_excl);
if (dma_fence_is_signaled(fence_excl))
dma_fence_put(fence_excl);
else
fence = fence_excl;
}
......@@ -425,12 +427,12 @@ long reservation_object_wait_timeout_rcu(struct reservation_object *obj,
rcu_read_unlock();
if (fence) {
if (read_seqcount_retry(&obj->seq, seq)) {
fence_put(fence);
dma_fence_put(fence);
goto retry;
}
ret = fence_wait_timeout(fence, intr, ret);
fence_put(fence);
ret = dma_fence_wait_timeout(fence, intr, ret);
dma_fence_put(fence);
if (ret > 0 && wait_all && (i + 1 < shared_count))
goto retry;
}
......@@ -444,18 +446,18 @@ EXPORT_SYMBOL_GPL(reservation_object_wait_timeout_rcu);
static inline int
reservation_object_test_signaled_single(struct fence *passed_fence)
reservation_object_test_signaled_single(struct dma_fence *passed_fence)
{
struct fence *fence, *lfence = passed_fence;
struct dma_fence *fence, *lfence = passed_fence;
int ret = 1;
if (!test_bit(FENCE_FLAG_SIGNALED_BIT, &lfence->flags)) {
fence = fence_get_rcu(lfence);
if (!test_bit(DMA_FENCE_FLAG_SIGNALED_BIT, &lfence->flags)) {
fence = dma_fence_get_rcu(lfence);
if (!fence)
return -1;
ret = !!fence_is_signaled(fence);
fence_put(fence);
ret = !!dma_fence_is_signaled(fence);
dma_fence_put(fence);
}
return ret;
}
......@@ -492,7 +494,7 @@ bool reservation_object_test_signaled_rcu(struct reservation_object *obj,
shared_count = fobj->shared_count;
for (i = 0; i < shared_count; ++i) {
struct fence *fence = rcu_dereference(fobj->shared[i]);
struct dma_fence *fence = rcu_dereference(fobj->shared[i]);
ret = reservation_object_test_signaled_single(fence);
if (ret < 0)
......@@ -506,7 +508,7 @@ bool reservation_object_test_signaled_rcu(struct reservation_object *obj,
}
if (!shared_count) {
struct fence *fence_excl = rcu_dereference(obj->fence_excl);
struct dma_fence *fence_excl = rcu_dereference(obj->fence_excl);
if (fence_excl) {
ret = reservation_object_test_signaled_single(
......
......@@ -21,35 +21,35 @@
#include <linux/export.h>
#include <linux/seqno-fence.h>
static const char *seqno_fence_get_driver_name(struct fence *fence)
static const char *seqno_fence_get_driver_name(struct dma_fence *fence)
{
struct seqno_fence *seqno_fence = to_seqno_fence(fence);
return seqno_fence->ops->get_driver_name(fence);
}
static const char *seqno_fence_get_timeline_name(struct fence *fence)
static const char *seqno_fence_get_timeline_name(struct dma_fence *fence)
{
struct seqno_fence *seqno_fence = to_seqno_fence(fence);
return seqno_fence->ops->get_timeline_name(fence);
}
static bool seqno_enable_signaling(struct fence *fence)
static bool seqno_enable_signaling(struct dma_fence *fence)
{
struct seqno_fence *seqno_fence = to_seqno_fence(fence);
return seqno_fence->ops->enable_signaling(fence);
}
static bool seqno_signaled(struct fence *fence)
static bool seqno_signaled(struct dma_fence *fence)
{
struct seqno_fence *seqno_fence = to_seqno_fence(fence);
return seqno_fence->ops->signaled && seqno_fence->ops->signaled(fence);
}
static void seqno_release(struct fence *fence)
static void seqno_release(struct dma_fence *fence)
{
struct seqno_fence *f = to_seqno_fence(fence);
......@@ -57,18 +57,18 @@ static void seqno_release(struct fence *fence)
if (f->ops->release)
f->ops->release(fence);
else
fence_free(&f->base);
dma_fence_free(&f->base);
}
static signed long seqno_wait(struct fence *fence, bool intr,
signed long timeout)
static signed long seqno_wait(struct dma_fence *fence, bool intr,
signed long timeout)
{
struct seqno_fence *f = to_seqno_fence(fence);
return f->ops->wait(fence, intr, timeout);
}
const struct fence_ops seqno_fence_ops = {
const struct dma_fence_ops seqno_fence_ops = {
.get_driver_name = seqno_fence_get_driver_name,
.get_timeline_name = seqno_fence_get_timeline_name,
.enable_signaling = seqno_enable_signaling,
......
......@@ -68,9 +68,9 @@ struct sw_sync_create_fence_data {
#define SW_SYNC_IOC_INC _IOW(SW_SYNC_IOC_MAGIC, 1, __u32)
static const struct fence_ops timeline_fence_ops;
static const struct dma_fence_ops timeline_fence_ops;
static inline struct sync_pt *fence_to_sync_pt(struct fence *fence)
static inline struct sync_pt *dma_fence_to_sync_pt(struct dma_fence *fence)
{
if (fence->ops != &timeline_fence_ops)
return NULL;
......@@ -93,7 +93,7 @@ struct sync_timeline *sync_timeline_create(const char *name)
return NULL;
kref_init(&obj->kref);
obj->context = fence_context_alloc(1);
obj->context = dma_fence_context_alloc(1);
strlcpy(obj->name, name, sizeof(obj->name));
INIT_LIST_HEAD(&obj->child_list_head);
......@@ -146,7 +146,7 @@ static void sync_timeline_signal(struct sync_timeline *obj, unsigned int inc)
list_for_each_entry_safe(pt, next, &obj->active_list_head,
active_list) {
if (fence_is_signaled_locked(&pt->base))
if (dma_fence_is_signaled_locked(&pt->base))
list_del_init(&pt->active_list);
}
......@@ -179,30 +179,30 @@ static struct sync_pt *sync_pt_create(struct sync_timeline *obj, int size,
spin_lock_irqsave(&obj->child_list_lock, flags);
sync_timeline_get(obj);
fence_init(&pt->base, &timeline_fence_ops, &obj->child_list_lock,
obj->context, value);
dma_fence_init(&pt->base, &timeline_fence_ops, &obj->child_list_lock,
obj->context, value);
list_add_tail(&pt->child_list, &obj->child_list_head);
INIT_LIST_HEAD(&pt->active_list);
spin_unlock_irqrestore(&obj->child_list_lock, flags);
return pt;
}
static const char *timeline_fence_get_driver_name(struct fence *fence)
static const char *timeline_fence_get_driver_name(struct dma_fence *fence)
{
return "sw_sync";
}
static const char *timeline_fence_get_timeline_name(struct fence *fence)
static const char *timeline_fence_get_timeline_name(struct dma_fence *fence)
{
struct sync_timeline *parent = fence_parent(fence);
struct sync_timeline *parent = dma_fence_parent(fence);
return parent->name;
}
static void timeline_fence_release(struct fence *fence)
static void timeline_fence_release(struct dma_fence *fence)
{
struct sync_pt *pt = fence_to_sync_pt(fence);
struct sync_timeline *parent = fence_parent(fence);
struct sync_pt *pt = dma_fence_to_sync_pt(fence);
struct sync_timeline *parent = dma_fence_parent(fence);
unsigned long flags;
spin_lock_irqsave(fence->lock, flags);
......@@ -212,20 +212,20 @@ static void timeline_fence_release(struct fence *fence)
spin_unlock_irqrestore(fence->lock, flags);
sync_timeline_put(parent);
fence_free(fence);
dma_fence_free(fence);
}
static bool timeline_fence_signaled(struct fence *fence)
static bool timeline_fence_signaled(struct dma_fence *fence)
{
struct sync_timeline *parent = fence_parent(fence);
struct sync_timeline *parent = dma_fence_parent(fence);
return (fence->seqno > parent->value) ? false : true;
}
static bool timeline_fence_enable_signaling(struct fence *fence)
static bool timeline_fence_enable_signaling(struct dma_fence *fence)
{
struct sync_pt *pt = fence_to_sync_pt(fence);
struct sync_timeline *parent = fence_parent(fence);
struct sync_pt *pt = dma_fence_to_sync_pt(fence);
struct sync_timeline *parent = dma_fence_parent(fence);
if (timeline_fence_signaled(fence))
return false;
......@@ -234,26 +234,26 @@ static bool timeline_fence_enable_signaling(struct fence *fence)
return true;
}
static void timeline_fence_value_str(struct fence *fence,
static void timeline_fence_value_str(struct dma_fence *fence,
char *str, int size)
{
snprintf(str, size, "%d", fence->seqno);
}
static void timeline_fence_timeline_value_str(struct fence *fence,
static void timeline_fence_timeline_value_str(struct dma_fence *fence,
char *str, int size)
{
struct sync_timeline *parent = fence_parent(fence);
struct sync_timeline *parent = dma_fence_parent(fence);
snprintf(str, size, "%d", parent->value);
}
static const struct fence_ops timeline_fence_ops = {
static const struct dma_fence_ops timeline_fence_ops = {
.get_driver_name = timeline_fence_get_driver_name,
.get_timeline_name = timeline_fence_get_timeline_name,
.enable_signaling = timeline_fence_enable_signaling,
.signaled = timeline_fence_signaled,
.wait = fence_default_wait,
.wait = dma_fence_default_wait,
.release = timeline_fence_release,
.fence_value_str = timeline_fence_value_str,
.timeline_value_str = timeline_fence_timeline_value_str,
......@@ -317,7 +317,7 @@ static long sw_sync_ioctl_create_fence(struct sync_timeline *obj,
sync_file = sync_file_create(&pt->base);
if (!sync_file) {
fence_put(&pt->base);
dma_fence_put(&pt->base);
err = -ENOMEM;
goto err;
}
......
......@@ -71,12 +71,13 @@ static const char *sync_status_str(int status)
return "error";
}
static void sync_print_fence(struct seq_file *s, struct fence *fence, bool show)
static void sync_print_fence(struct seq_file *s,
struct dma_fence *fence, bool show)
{
int status = 1;
struct sync_timeline *parent = fence_parent(fence);
struct sync_timeline *parent = dma_fence_parent(fence);
if (fence_is_signaled_locked(fence))
if (dma_fence_is_signaled_locked(fence))
status = fence->status;
seq_printf(s, " %s%sfence %s",
......@@ -135,10 +136,10 @@ static void sync_print_sync_file(struct seq_file *s,
int i;
seq_printf(s, "[%p] %s: %s\n", sync_file, sync_file->name,
sync_status_str(!fence_is_signaled(sync_file->fence)));
sync_status_str(!dma_fence_is_signaled(sync_file->fence)));
if (fence_is_array(sync_file->fence)) {
struct fence_array *array = to_fence_array(sync_file->fence);
if (dma_fence_is_array(sync_file->fence)) {
struct dma_fence_array *array = to_dma_fence_array(sync_file->fence);
for (i = 0; i < array->num_fences; ++i)
sync_print_fence(s, array->fences[i], true);
......
......@@ -15,7 +15,7 @@
#include <linux/list.h>
#include <linux/spinlock.h>
#include <linux/fence.h>
#include <linux/dma-fence.h>
#include <linux/sync_file.h>
#include <uapi/linux/sync_file.h>
......@@ -45,10 +45,9 @@ struct sync_timeline {
struct list_head sync_timeline_list;
};
static inline struct sync_timeline *fence_parent(struct fence *fence)
static inline struct sync_timeline *dma_fence_parent(struct dma_fence *fence)
{
return container_of(fence->lock, struct sync_timeline,
child_list_lock);
return container_of(fence->lock, struct sync_timeline, child_list_lock);
}
/**
......@@ -58,7 +57,7 @@ static inline struct sync_timeline *fence_parent(struct fence *fence)
* @active_list: sync timeline active child's list
*/
struct sync_pt {
struct fence base;
struct dma_fence base;
struct list_head child_list;
struct list_head active_list;
};
......
......@@ -54,7 +54,7 @@ static struct sync_file *sync_file_alloc(void)
return NULL;
}
static void fence_check_cb_func(struct fence *f, struct fence_cb *cb)
static void fence_check_cb_func(struct dma_fence *f, struct dma_fence_cb *cb)
{
struct sync_file *sync_file;
......@@ -71,7 +71,7 @@ static void fence_check_cb_func(struct fence *f, struct fence_cb *cb)
* takes ownership of @fence. The sync_file can be released with
* fput(sync_file->file). Returns the sync_file or NULL in case of error.
*/
struct sync_file *sync_file_create(struct fence *fence)
struct sync_file *sync_file_create(struct dma_fence *fence)
{
struct sync_file *sync_file;
......@@ -79,7 +79,7 @@ struct sync_file *sync_file_create(struct fence *fence)
if (!sync_file)
return NULL;
sync_file->fence = fence_get(fence);
sync_file->fence = dma_fence_get(fence);
snprintf(sync_file->name, sizeof(sync_file->name), "%s-%s%llu-%d",
fence->ops->get_driver_name(fence),
......@@ -121,16 +121,16 @@ static struct sync_file *sync_file_fdget(int fd)
* Ensures @fd references a valid sync_file and returns a fence that
* represents all fence in the sync_file. On error NULL is returned.
*/
struct fence *sync_file_get_fence(int fd)
struct dma_fence *sync_file_get_fence(int fd)
{
struct sync_file *sync_file;
struct fence *fence;
struct dma_fence *fence;
sync_file = sync_file_fdget(fd);
if (!sync_file)
return NULL;
fence = fence_get(sync_file->fence);
fence = dma_fence_get(sync_file->fence);
fput(sync_file->file);
return fence;
......@@ -138,22 +138,23 @@ struct fence *sync_file_get_fence(int fd)
EXPORT_SYMBOL(sync_file_get_fence);
static int sync_file_set_fence(struct sync_file *sync_file,
struct fence **fences, int num_fences)
struct dma_fence **fences, int num_fences)
{
struct fence_array *array;
struct dma_fence_array *array;
/*
* The reference for the fences in the new sync_file and held
* in add_fence() during the merge procedure, so for num_fences == 1
* we already own a new reference to the fence. For num_fence > 1
* we own the reference of the fence_array creation.
* we own the reference of the dma_fence_array creation.
*/
if (num_fences == 1) {
sync_file->fence = fences[0];
kfree(fences);
} else {
array = fence_array_create(num_fences, fences,
fence_context_alloc(1), 1, false);
array = dma_fence_array_create(num_fences, fences,
dma_fence_context_alloc(1),
1, false);
if (!array)
return -ENOMEM;
......@@ -163,10 +164,11 @@ static int sync_file_set_fence(struct sync_file *sync_file,
return 0;
}
static struct fence **get_fences(struct sync_file *sync_file, int *num_fences)
static struct dma_fence **get_fences(struct sync_file *sync_file,
int *num_fences)
{
if (fence_is_array(sync_file->fence)) {
struct fence_array *array = to_fence_array(sync_file->fence);
if (dma_fence_is_array(sync_file->fence)) {
struct dma_fence_array *array = to_dma_fence_array(sync_file->fence);
*num_fences = array->num_fences;
return array->fences;
......@@ -176,12 +178,13 @@ static struct fence **get_fences(struct sync_file *sync_file, int *num_fences)
return &sync_file->fence;
}
static void add_fence(struct fence **fences, int *i, struct fence *fence)
static void add_fence(struct dma_fence **fences,
int *i, struct dma_fence *fence)
{
fences[*i] = fence;
if (!fence_is_signaled(fence)) {
fence_get(fence);
if (!dma_fence_is_signaled(fence)) {
dma_fence_get(fence);
(*i)++;
}
}
......@@ -200,7 +203,7 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a,
struct sync_file *b)
{
struct sync_file *sync_file;
struct fence **fences, **nfences, **a_fences, **b_fences;
struct dma_fence **fences, **nfences, **a_fences, **b_fences;
int i, i_a, i_b, num_fences, a_num_fences, b_num_fences;
sync_file = sync_file_alloc();
......@@ -226,8 +229,8 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a,
* and sync_file_create, this is a reasonable assumption.
*/
for (i = i_a = i_b = 0; i_a < a_num_fences && i_b < b_num_fences; ) {
struct fence *pt_a = a_fences[i_a];
struct fence *pt_b = b_fences[i_b];
struct dma_fence *pt_a = a_fences[i_a];
struct dma_fence *pt_b = b_fences[i_b];
if (pt_a->context < pt_b->context) {
add_fence(fences, &i, pt_a);
......@@ -255,7 +258,7 @@ static struct sync_file *sync_file_merge(const char *name, struct sync_file *a,
add_fence(fences, &i, b_fences[i_b]);
if (i == 0)
fences[i++] = fence_get(a_fences[0]);
fences[i++] = dma_fence_get(a_fences[0]);
if (num_fences > i) {
nfences = krealloc(fences, i * sizeof(*fences),
......@@ -286,8 +289,8 @@ static void sync_file_free(struct kref *kref)
kref);
if (test_bit(POLL_ENABLED, &sync_file->fence->flags))
fence_remove_callback(sync_file->fence, &sync_file->cb);
fence_put(sync_file->fence);
dma_fence_remove_callback(sync_file->fence, &sync_file->cb);
dma_fence_put(sync_file->fence);
kfree(sync_file);
}
......@@ -307,12 +310,12 @@ static unsigned int sync_file_poll(struct file *file, poll_table *wait)
if (!poll_does_not_wait(wait) &&
!test_and_set_bit(POLL_ENABLED, &sync_file->fence->flags)) {
if (fence_add_callback(sync_file->fence, &sync_file->cb,
fence_check_cb_func) < 0)
if (dma_fence_add_callback(sync_file->fence, &sync_file->cb,
fence_check_cb_func) < 0)
wake_up_all(&sync_file->wq);
}
return fence_is_signaled(sync_file->fence) ? POLLIN : 0;
return dma_fence_is_signaled(sync_file->fence) ? POLLIN : 0;
}
static long sync_file_ioctl_merge(struct sync_file *sync_file,
......@@ -370,14 +373,14 @@ static long sync_file_ioctl_merge(struct sync_file *sync_file,
return err;
}
static void sync_fill_fence_info(struct fence *fence,
static void sync_fill_fence_info(struct dma_fence *fence,
struct sync_fence_info *info)
{
strlcpy(info->obj_name, fence->ops->get_timeline_name(fence),
sizeof(info->obj_name));
strlcpy(info->driver_name, fence->ops->get_driver_name(fence),
sizeof(info->driver_name));
if (fence_is_signaled(fence))
if (dma_fence_is_signaled(fence))
info->status = fence->status >= 0 ? 1 : fence->status;
else
info->status = 0;
......@@ -389,7 +392,7 @@ static long sync_file_ioctl_fence_info(struct sync_file *sync_file,
{
struct sync_file_info info;
struct sync_fence_info *fence_info = NULL;
struct fence **fences;
struct dma_fence **fences;
__u32 size;
int num_fences, ret, i;
......@@ -429,7 +432,7 @@ static long sync_file_ioctl_fence_info(struct sync_file *sync_file,
no_fences:
strlcpy(info.name, sync_file->name, sizeof(info.name));
info.status = fence_is_signaled(sync_file->fence);
info.status = dma_fence_is_signaled(sync_file->fence);
info.num_fences = num_fences;
if (copy_to_user((void __user *)arg, &info, sizeof(info)))
......
......@@ -24,7 +24,7 @@ amdgpu-y += amdgpu_device.o amdgpu_kms.o \
atombios_encoders.o amdgpu_sa.o atombios_i2c.o \
amdgpu_prime.o amdgpu_vm.o amdgpu_ib.o amdgpu_pll.o \
amdgpu_ucode.o amdgpu_bo_list.o amdgpu_ctx.o amdgpu_sync.o \
amdgpu_gtt_mgr.o
amdgpu_gtt_mgr.o amdgpu_vram_mgr.o
# add asic specific block
amdgpu-$(CONFIG_DRM_AMDGPU_CIK)+= cik.o cik_ih.o kv_smc.o kv_dpm.o \
......
......@@ -90,7 +90,6 @@
#define ENCODER_OBJECT_ID_INTERNAL_VCE 0x24
#define ENCODER_OBJECT_ID_INTERNAL_UNIPHY3 0x25
#define ENCODER_OBJECT_ID_INTERNAL_AMCLK 0x27
#define ENCODER_OBJECT_ID_VIRTUAL 0x28
#define ENCODER_OBJECT_ID_GENERAL_EXTERNAL_DVO 0xFF
......@@ -120,7 +119,6 @@
#define CONNECTOR_OBJECT_ID_eDP 0x14
#define CONNECTOR_OBJECT_ID_MXM 0x15
#define CONNECTOR_OBJECT_ID_LVDS_eDP 0x16
#define CONNECTOR_OBJECT_ID_VIRTUAL 0x17
/* deleted */
......@@ -149,7 +147,6 @@
#define GRAPH_OBJECT_ENUM_ID5 0x05
#define GRAPH_OBJECT_ENUM_ID6 0x06
#define GRAPH_OBJECT_ENUM_ID7 0x07
#define GRAPH_OBJECT_ENUM_VIRTUAL 0x08
/****************************************************/
/* Graphics Object ID Bit definition */
......@@ -411,10 +408,6 @@
GRAPH_OBJECT_ENUM_ID1 << ENUM_ID_SHIFT |\
ENCODER_OBJECT_ID_HDMI_ANX9805 << OBJECT_ID_SHIFT)
#define ENCODER_VIRTUAL_ENUM_VIRTUAL ( GRAPH_OBJECT_TYPE_ENCODER << OBJECT_TYPE_SHIFT |\
GRAPH_OBJECT_ENUM_VIRTUAL << ENUM_ID_SHIFT |\
ENCODER_OBJECT_ID_VIRTUAL << OBJECT_ID_SHIFT)
/****************************************************/
/* Connector Object ID definition - Shared with BIOS */
/****************************************************/
......
此差异已折叠。
......@@ -265,14 +265,14 @@ static int acp_hw_init(void *handle)
struct amdgpu_device *adev = (struct amdgpu_device *)handle;
const struct amdgpu_ip_block_version *ip_version =
const struct amdgpu_ip_block *ip_block =
amdgpu_get_ip_block(adev, AMD_IP_BLOCK_TYPE_ACP);
if (!ip_version)
if (!ip_block)
return -EINVAL;
r = amd_acp_hw_init(adev->acp.cgs_device,
ip_version->major, ip_version->minor);
ip_block->version->major, ip_block->version->minor);
/* -ENODEV means board uses AZ rather than ACP */
if (r == -ENODEV)
return 0;
......@@ -456,7 +456,7 @@ static int acp_set_powergating_state(void *handle,
return 0;
}
const struct amd_ip_funcs acp_ip_funcs = {
static const struct amd_ip_funcs acp_ip_funcs = {
.name = "acp_ip",
.early_init = acp_early_init,
.late_init = NULL,
......@@ -472,3 +472,12 @@ const struct amd_ip_funcs acp_ip_funcs = {
.set_clockgating_state = acp_set_clockgating_state,
.set_powergating_state = acp_set_powergating_state,
};
const struct amdgpu_ip_block_version acp_ip_block =
{
.type = AMD_IP_BLOCK_TYPE_ACP,
.major = 2,
.minor = 2,
.rev = 0,
.funcs = &acp_ip_funcs,
};
......@@ -37,6 +37,6 @@ struct amdgpu_acp {
struct acp_pm_domain *acp_genpd;
};
extern const struct amd_ip_funcs acp_ip_funcs;
extern const struct amdgpu_ip_block_version acp_ip_block;
#endif /* __AMDGPU_ACP_H__ */
......@@ -1115,49 +1115,6 @@ int amdgpu_atombios_get_memory_pll_dividers(struct amdgpu_device *adev,
return 0;
}
uint32_t amdgpu_atombios_get_engine_clock(struct amdgpu_device *adev)
{
GET_ENGINE_CLOCK_PS_ALLOCATION args;
int index = GetIndexIntoMasterTable(COMMAND, GetEngineClock);
amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
return le32_to_cpu(args.ulReturnEngineClock);
}
uint32_t amdgpu_atombios_get_memory_clock(struct amdgpu_device *adev)
{
GET_MEMORY_CLOCK_PS_ALLOCATION args;
int index = GetIndexIntoMasterTable(COMMAND, GetMemoryClock);
amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
return le32_to_cpu(args.ulReturnMemoryClock);
}
void amdgpu_atombios_set_engine_clock(struct amdgpu_device *adev,
uint32_t eng_clock)
{
SET_ENGINE_CLOCK_PS_ALLOCATION args;
int index = GetIndexIntoMasterTable(COMMAND, SetEngineClock);
args.ulTargetEngineClock = cpu_to_le32(eng_clock); /* 10 khz */
amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
}
void amdgpu_atombios_set_memory_clock(struct amdgpu_device *adev,
uint32_t mem_clock)
{
SET_MEMORY_CLOCK_PS_ALLOCATION args;
int index = GetIndexIntoMasterTable(COMMAND, SetMemoryClock);
if (adev->flags & AMD_IS_APU)
return;
args.ulTargetMemoryClock = cpu_to_le32(mem_clock); /* 10 khz */
amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
}
void amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev,
u32 eng_clock, u32 mem_clock)
{
......@@ -1256,45 +1213,6 @@ int amdgpu_atombios_get_leakage_vddc_based_on_leakage_idx(struct amdgpu_device *
return amdgpu_atombios_get_max_vddc(adev, VOLTAGE_TYPE_VDDC, leakage_idx, voltage);
}
void amdgpu_atombios_set_voltage(struct amdgpu_device *adev,
u16 voltage_level,
u8 voltage_type)
{
union set_voltage args;
int index = GetIndexIntoMasterTable(COMMAND, SetVoltage);
u8 frev, crev, volt_index = voltage_level;
if (!amdgpu_atom_parse_cmd_header(adev->mode_info.atom_context, index, &frev, &crev))
return;
/* 0xff01 is a flag rather then an actual voltage */
if (voltage_level == 0xff01)
return;
switch (crev) {
case 1:
args.v1.ucVoltageType = voltage_type;
args.v1.ucVoltageMode = SET_ASIC_VOLTAGE_MODE_ALL_SOURCE;
args.v1.ucVoltageIndex = volt_index;
break;
case 2:
args.v2.ucVoltageType = voltage_type;
args.v2.ucVoltageMode = SET_ASIC_VOLTAGE_MODE_SET_VOLTAGE;
args.v2.usVoltageLevel = cpu_to_le16(voltage_level);
break;
case 3:
args.v3.ucVoltageType = voltage_type;
args.v3.ucVoltageMode = ATOM_SET_VOLTAGE;
args.v3.usVoltageLevel = cpu_to_le16(voltage_level);
break;
default:
DRM_ERROR("Unknown table version %d, %d\n", frev, crev);
return;
}
amdgpu_atom_execute_table(adev->mode_info.atom_context, index, (uint32_t *)&args);
}
int amdgpu_atombios_get_leakage_id_from_vbios(struct amdgpu_device *adev,
u16 *leakage_id)
{
......@@ -1784,6 +1702,19 @@ void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev)
WREG32(mmBIOS_SCRATCH_0 + i, adev->bios_scratch[i]);
}
void amdgpu_atombios_scratch_regs_engine_hung(struct amdgpu_device *adev,
bool hung)
{
u32 tmp = RREG32(mmBIOS_SCRATCH_3);
if (hung)
tmp |= ATOM_S3_ASIC_GUI_ENGINE_HUNG;
else
tmp &= ~ATOM_S3_ASIC_GUI_ENGINE_HUNG;
WREG32(mmBIOS_SCRATCH_3, tmp);
}
/* Atom needs data in little endian format
* so swap as appropriate when copying data to
* or from atom. Note that atom operates on
......
......@@ -163,16 +163,6 @@ int amdgpu_atombios_get_memory_pll_dividers(struct amdgpu_device *adev,
bool strobe_mode,
struct atom_mpll_param *mpll_param);
uint32_t amdgpu_atombios_get_engine_clock(struct amdgpu_device *adev);
uint32_t amdgpu_atombios_get_memory_clock(struct amdgpu_device *adev);
void amdgpu_atombios_set_engine_clock(struct amdgpu_device *adev,
uint32_t eng_clock);
void amdgpu_atombios_set_memory_clock(struct amdgpu_device *adev,
uint32_t mem_clock);
void amdgpu_atombios_set_voltage(struct amdgpu_device *adev,
u16 voltage_level,
u8 voltage_type);
void amdgpu_atombios_set_engine_dram_timings(struct amdgpu_device *adev,
u32 eng_clock, u32 mem_clock);
......@@ -206,6 +196,8 @@ void amdgpu_atombios_scratch_regs_lock(struct amdgpu_device *adev, bool lock);
void amdgpu_atombios_scratch_regs_init(struct amdgpu_device *adev);
void amdgpu_atombios_scratch_regs_save(struct amdgpu_device *adev);
void amdgpu_atombios_scratch_regs_restore(struct amdgpu_device *adev);
void amdgpu_atombios_scratch_regs_engine_hung(struct amdgpu_device *adev,
bool hung);
void amdgpu_atombios_copy_swap(u8 *dst, u8 *src, u8 num_bytes, bool to_le);
int amdgpu_atombios_get_max_vddc(struct amdgpu_device *adev, u8 voltage_type,
......
......@@ -33,7 +33,7 @@ static int amdgpu_benchmark_do_move(struct amdgpu_device *adev, unsigned size,
{
unsigned long start_jiffies;
unsigned long end_jiffies;
struct fence *fence = NULL;
struct dma_fence *fence = NULL;
int i, r;
start_jiffies = jiffies;
......@@ -43,17 +43,17 @@ static int amdgpu_benchmark_do_move(struct amdgpu_device *adev, unsigned size,
false);
if (r)
goto exit_do_move;
r = fence_wait(fence, false);
r = dma_fence_wait(fence, false);
if (r)
goto exit_do_move;
fence_put(fence);
dma_fence_put(fence);
}
end_jiffies = jiffies;
r = jiffies_to_msecs(end_jiffies - start_jiffies);
exit_do_move:
if (fence)
fence_put(fence);
dma_fence_put(fence);
return r;
}
......
......@@ -146,7 +146,8 @@ static int amdgpu_cgs_alloc_gpu_mem(struct cgs_device *cgs_device,
switch(type) {
case CGS_GPU_MEM_TYPE__VISIBLE_CONTIG_FB:
case CGS_GPU_MEM_TYPE__VISIBLE_FB:
flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED;
flags = AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
domain = AMDGPU_GEM_DOMAIN_VRAM;
if (max_offset > adev->mc.real_vram_size)
return -EINVAL;
......@@ -157,7 +158,8 @@ static int amdgpu_cgs_alloc_gpu_mem(struct cgs_device *cgs_device,
break;
case CGS_GPU_MEM_TYPE__INVISIBLE_CONTIG_FB:
case CGS_GPU_MEM_TYPE__INVISIBLE_FB:
flags = AMDGPU_GEM_CREATE_NO_CPU_ACCESS;
flags = AMDGPU_GEM_CREATE_NO_CPU_ACCESS |
AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
domain = AMDGPU_GEM_DOMAIN_VRAM;
if (adev->mc.visible_vram_size < adev->mc.real_vram_size) {
place.fpfn =
......@@ -240,7 +242,7 @@ static int amdgpu_cgs_gmap_gpu_mem(struct cgs_device *cgs_device, cgs_handle_t h
r = amdgpu_bo_reserve(obj, false);
if (unlikely(r != 0))
return r;
r = amdgpu_bo_pin_restricted(obj, AMDGPU_GEM_DOMAIN_GTT,
r = amdgpu_bo_pin_restricted(obj, obj->prefered_domains,
min_offset, max_offset, mcaddr);
amdgpu_bo_unreserve(obj);
return r;
......@@ -624,11 +626,11 @@ static int amdgpu_cgs_set_clockgating_state(struct cgs_device *cgs_device,
int i, r = -1;
for (i = 0; i < adev->num_ip_blocks; i++) {
if (!adev->ip_block_status[i].valid)
if (!adev->ip_blocks[i].status.valid)
continue;
if (adev->ip_blocks[i].type == block_type) {
r = adev->ip_blocks[i].funcs->set_clockgating_state(
if (adev->ip_blocks[i].version->type == block_type) {
r = adev->ip_blocks[i].version->funcs->set_clockgating_state(
(void *)adev,
state);
break;
......@@ -645,11 +647,11 @@ static int amdgpu_cgs_set_powergating_state(struct cgs_device *cgs_device,
int i, r = -1;
for (i = 0; i < adev->num_ip_blocks; i++) {
if (!adev->ip_block_status[i].valid)
if (!adev->ip_blocks[i].status.valid)
continue;
if (adev->ip_blocks[i].type == block_type) {
r = adev->ip_blocks[i].funcs->set_powergating_state(
if (adev->ip_blocks[i].version->type == block_type) {
r = adev->ip_blocks[i].version->funcs->set_powergating_state(
(void *)adev,
state);
break;
......@@ -685,15 +687,21 @@ static uint32_t fw_type_convert(struct cgs_device *cgs_device, uint32_t fw_type)
result = AMDGPU_UCODE_ID_CP_MEC1;
break;
case CGS_UCODE_ID_CP_MEC_JT2:
if (adev->asic_type == CHIP_TONGA || adev->asic_type == CHIP_POLARIS11
|| adev->asic_type == CHIP_POLARIS10)
result = AMDGPU_UCODE_ID_CP_MEC2;
else
/* for VI. JT2 should be the same as JT1, because:
1, MEC2 and MEC1 use exactly same FW.
2, JT2 is not pached but JT1 is.
*/
if (adev->asic_type >= CHIP_TOPAZ)
result = AMDGPU_UCODE_ID_CP_MEC1;
else
result = AMDGPU_UCODE_ID_CP_MEC2;
break;
case CGS_UCODE_ID_RLC_G:
result = AMDGPU_UCODE_ID_RLC_G;
break;
case CGS_UCODE_ID_STORAGE:
result = AMDGPU_UCODE_ID_STORAGE;
break;
default:
DRM_ERROR("Firmware type not supported\n");
}
......@@ -776,12 +784,18 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
if ((type == CGS_UCODE_ID_CP_MEC_JT1) ||
(type == CGS_UCODE_ID_CP_MEC_JT2)) {
gpu_addr += le32_to_cpu(header->jt_offset) << 2;
gpu_addr += ALIGN(le32_to_cpu(header->header.ucode_size_bytes), PAGE_SIZE);
data_size = le32_to_cpu(header->jt_size) << 2;
}
info->mc_addr = gpu_addr;
info->kptr = ucode->kaddr;
info->image_size = data_size;
info->mc_addr = gpu_addr;
info->version = (uint16_t)le32_to_cpu(header->header.ucode_version);
if (CGS_UCODE_ID_CP_MEC == type)
info->image_size = (header->jt_offset) << 2;
info->fw_version = amdgpu_get_firmware_version(cgs_device, type);
info->feature_version = (uint16_t)le32_to_cpu(header->ucode_feature_version);
} else {
......@@ -851,6 +865,12 @@ static int amdgpu_cgs_get_firmware_info(struct cgs_device *cgs_device,
return 0;
}
static int amdgpu_cgs_is_virtualization_enabled(void *cgs_device)
{
CGS_FUNC_ADEV;
return amdgpu_sriov_vf(adev);
}
static int amdgpu_cgs_query_system_info(struct cgs_device *cgs_device,
struct cgs_system_info *sys_info)
{
......@@ -1204,6 +1224,7 @@ static const struct cgs_ops amdgpu_cgs_ops = {
amdgpu_cgs_notify_dpm_enabled,
amdgpu_cgs_call_acpi_method,
amdgpu_cgs_query_system_info,
amdgpu_cgs_is_virtualization_enabled
};
static const struct cgs_os_ops amdgpu_cgs_os_ops = {
......
......@@ -1517,88 +1517,6 @@ static const struct drm_connector_funcs amdgpu_connector_edp_funcs = {
.force = amdgpu_connector_dvi_force,
};
static struct drm_encoder *
amdgpu_connector_virtual_encoder(struct drm_connector *connector)
{
int enc_id = connector->encoder_ids[0];
struct drm_encoder *encoder;
int i;
for (i = 0; i < DRM_CONNECTOR_MAX_ENCODER; i++) {
if (connector->encoder_ids[i] == 0)
break;
encoder = drm_encoder_find(connector->dev, connector->encoder_ids[i]);
if (!encoder)
continue;
if (encoder->encoder_type == DRM_MODE_ENCODER_VIRTUAL)
return encoder;
}
/* pick the first one */
if (enc_id)
return drm_encoder_find(connector->dev, enc_id);
return NULL;
}
static int amdgpu_connector_virtual_get_modes(struct drm_connector *connector)
{
struct drm_encoder *encoder = amdgpu_connector_best_single_encoder(connector);
if (encoder) {
amdgpu_connector_add_common_modes(encoder, connector);
}
return 0;
}
static int amdgpu_connector_virtual_mode_valid(struct drm_connector *connector,
struct drm_display_mode *mode)
{
return MODE_OK;
}
static int
amdgpu_connector_virtual_dpms(struct drm_connector *connector, int mode)
{
return 0;
}
static enum drm_connector_status
amdgpu_connector_virtual_detect(struct drm_connector *connector, bool force)
{
return connector_status_connected;
}
static int
amdgpu_connector_virtual_set_property(struct drm_connector *connector,
struct drm_property *property,
uint64_t val)
{
return 0;
}
static void amdgpu_connector_virtual_force(struct drm_connector *connector)
{
return;
}
static const struct drm_connector_helper_funcs amdgpu_connector_virtual_helper_funcs = {
.get_modes = amdgpu_connector_virtual_get_modes,
.mode_valid = amdgpu_connector_virtual_mode_valid,
.best_encoder = amdgpu_connector_virtual_encoder,
};
static const struct drm_connector_funcs amdgpu_connector_virtual_funcs = {
.dpms = amdgpu_connector_virtual_dpms,
.detect = amdgpu_connector_virtual_detect,
.fill_modes = drm_helper_probe_single_connector_modes,
.set_property = amdgpu_connector_virtual_set_property,
.destroy = amdgpu_connector_destroy,
.force = amdgpu_connector_virtual_force,
};
void
amdgpu_connector_add(struct amdgpu_device *adev,
uint32_t connector_id,
......@@ -1983,17 +1901,6 @@ amdgpu_connector_add(struct amdgpu_device *adev,
connector->interlace_allowed = false;
connector->doublescan_allowed = false;
break;
case DRM_MODE_CONNECTOR_VIRTUAL:
amdgpu_dig_connector = kzalloc(sizeof(struct amdgpu_connector_atom_dig), GFP_KERNEL);
if (!amdgpu_dig_connector)
goto failed;
amdgpu_connector->con_priv = amdgpu_dig_connector;
drm_connector_init(dev, &amdgpu_connector->base, &amdgpu_connector_virtual_funcs, connector_type);
drm_connector_helper_add(&amdgpu_connector->base, &amdgpu_connector_virtual_helper_funcs);
subpixel_order = SubPixelHorizontalRGB;
connector->interlace_allowed = false;
connector->doublescan_allowed = false;
break;
}
}
......
......@@ -355,6 +355,7 @@ static void amdgpu_cs_report_moved_bytes(struct amdgpu_device *adev,
static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
struct amdgpu_bo *bo)
{
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
u64 initial_bytes_moved;
uint32_t domain;
int r;
......@@ -372,9 +373,9 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
retry:
amdgpu_ttm_placement_from_domain(bo, domain);
initial_bytes_moved = atomic64_read(&bo->adev->num_bytes_moved);
initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) -
p->bytes_moved += atomic64_read(&adev->num_bytes_moved) -
initial_bytes_moved;
if (unlikely(r == -ENOMEM) && domain != bo->allowed_domains) {
......@@ -387,9 +388,9 @@ static int amdgpu_cs_bo_validate(struct amdgpu_cs_parser *p,
/* Last resort, try to evict something from the current working set */
static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
struct amdgpu_bo_list_entry *lobj)
struct amdgpu_bo *validated)
{
uint32_t domain = lobj->robj->allowed_domains;
uint32_t domain = validated->allowed_domains;
int r;
if (!p->evictable)
......@@ -400,11 +401,12 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
struct amdgpu_bo_list_entry *candidate = p->evictable;
struct amdgpu_bo *bo = candidate->robj;
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
u64 initial_bytes_moved;
uint32_t other;
/* If we reached our current BO we can forget it */
if (candidate == lobj)
if (candidate->robj == validated)
break;
other = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
......@@ -420,9 +422,9 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
/* Good we can try to move this BO somewhere else */
amdgpu_ttm_placement_from_domain(bo, other);
initial_bytes_moved = atomic64_read(&bo->adev->num_bytes_moved);
initial_bytes_moved = atomic64_read(&adev->num_bytes_moved);
r = ttm_bo_validate(&bo->tbo, &bo->placement, true, false);
p->bytes_moved += atomic64_read(&bo->adev->num_bytes_moved) -
p->bytes_moved += atomic64_read(&adev->num_bytes_moved) -
initial_bytes_moved;
if (unlikely(r))
......@@ -437,6 +439,23 @@ static bool amdgpu_cs_try_evict(struct amdgpu_cs_parser *p,
return false;
}
static int amdgpu_cs_validate(void *param, struct amdgpu_bo *bo)
{
struct amdgpu_cs_parser *p = param;
int r;
do {
r = amdgpu_cs_bo_validate(p, bo);
} while (r == -ENOMEM && amdgpu_cs_try_evict(p, bo));
if (r)
return r;
if (bo->shadow)
r = amdgpu_cs_bo_validate(p, bo);
return r;
}
static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
struct list_head *validated)
{
......@@ -464,18 +483,10 @@ static int amdgpu_cs_list_validate(struct amdgpu_cs_parser *p,
if (p->evictable == lobj)
p->evictable = NULL;
do {
r = amdgpu_cs_bo_validate(p, bo);
} while (r == -ENOMEM && amdgpu_cs_try_evict(p, lobj));
r = amdgpu_cs_validate(p, bo);
if (r)
return r;
if (bo->shadow) {
r = amdgpu_cs_bo_validate(p, bo);
if (r)
return r;
}
if (binding_userptr) {
drm_free_large(lobj->user_pages);
lobj->user_pages = NULL;
......@@ -593,14 +604,19 @@ static int amdgpu_cs_parser_bos(struct amdgpu_cs_parser *p,
list_splice(&need_pages, &p->validated);
}
amdgpu_vm_get_pt_bos(p->adev, &fpriv->vm, &duplicates);
p->bytes_moved_threshold = amdgpu_cs_get_threshold_for_moves(p->adev);
p->bytes_moved = 0;
p->evictable = list_last_entry(&p->validated,
struct amdgpu_bo_list_entry,
tv.head);
r = amdgpu_vm_validate_pt_bos(p->adev, &fpriv->vm,
amdgpu_cs_validate, p);
if (r) {
DRM_ERROR("amdgpu_vm_validate_pt_bos() failed.\n");
goto error_validate;
}
r = amdgpu_cs_list_validate(p, &duplicates);
if (r) {
DRM_ERROR("amdgpu_cs_list_validate(duplicates) failed.\n");
......@@ -719,7 +735,7 @@ static void amdgpu_cs_parser_fini(struct amdgpu_cs_parser *parser, int error, bo
ttm_eu_backoff_reservation(&parser->ticket,
&parser->validated);
}
fence_put(parser->fence);
dma_fence_put(parser->fence);
if (parser->ctx)
amdgpu_ctx_put(parser->ctx);
......@@ -756,7 +772,7 @@ static int amdgpu_bo_vm_update_pte(struct amdgpu_cs_parser *p,
if (p->bo_list) {
for (i = 0; i < p->bo_list->num_entries; i++) {
struct fence *f;
struct dma_fence *f;
/* ignore duplicates */
bo = p->bo_list->array[i].robj;
......@@ -806,13 +822,14 @@ static int amdgpu_cs_ib_vm_chunk(struct amdgpu_device *adev,
/* Only for UVD/VCE VM emulation */
if (ring->funcs->parse_cs) {
p->job->vm = NULL;
for (i = 0; i < p->job->num_ibs; i++) {
r = amdgpu_ring_parse_cs(ring, p, i);
if (r)
return r;
}
} else {
}
if (p->job->vm) {
p->job->vm_pd_addr = amdgpu_bo_gpu_offset(vm->page_directory);
r = amdgpu_bo_vm_update_pte(p, vm);
......@@ -901,7 +918,7 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
offset = ((uint64_t)m->it.start) * AMDGPU_GPU_PAGE_SIZE;
kptr += chunk_ib->va_start - offset;
r = amdgpu_ib_get(adev, NULL, chunk_ib->ib_bytes, ib);
r = amdgpu_ib_get(adev, vm, chunk_ib->ib_bytes, ib);
if (r) {
DRM_ERROR("Failed to get ib !\n");
return r;
......@@ -916,9 +933,9 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
return r;
}
ib->gpu_addr = chunk_ib->va_start;
}
ib->gpu_addr = chunk_ib->va_start;
ib->length_dw = chunk_ib->ib_bytes / 4;
ib->flags = chunk_ib->flags;
j++;
......@@ -926,8 +943,8 @@ static int amdgpu_cs_ib_fill(struct amdgpu_device *adev,
/* UVD & VCE fw doesn't support user fences */
if (parser->job->uf_addr && (
parser->job->ring->type == AMDGPU_RING_TYPE_UVD ||
parser->job->ring->type == AMDGPU_RING_TYPE_VCE))
parser->job->ring->funcs->type == AMDGPU_RING_TYPE_UVD ||
parser->job->ring->funcs->type == AMDGPU_RING_TYPE_VCE))
return -EINVAL;
return 0;
......@@ -956,7 +973,7 @@ static int amdgpu_cs_dependencies(struct amdgpu_device *adev,
for (j = 0; j < num_deps; ++j) {
struct amdgpu_ring *ring;
struct amdgpu_ctx *ctx;
struct fence *fence;
struct dma_fence *fence;
r = amdgpu_cs_get_ring(adev, deps[j].ip_type,
deps[j].ip_instance,
......@@ -978,7 +995,7 @@ static int amdgpu_cs_dependencies(struct amdgpu_device *adev,
} else if (fence) {
r = amdgpu_sync_fence(adev, &p->job->sync,
fence);
fence_put(fence);
dma_fence_put(fence);
amdgpu_ctx_put(ctx);
if (r)
return r;
......@@ -1008,7 +1025,7 @@ static int amdgpu_cs_submit(struct amdgpu_cs_parser *p,
job->owner = p->filp;
job->fence_ctx = entity->fence_context;
p->fence = fence_get(&job->base.s_fence->finished);
p->fence = dma_fence_get(&job->base.s_fence->finished);
cs->out.handle = amdgpu_ctx_add_fence(p->ctx, ring, p->fence);
job->uf_sequence = cs->out.handle;
amdgpu_job_free_resources(job);
......@@ -1091,7 +1108,7 @@ int amdgpu_cs_wait_ioctl(struct drm_device *dev, void *data,
unsigned long timeout = amdgpu_gem_timeout(wait->in.timeout);
struct amdgpu_ring *ring = NULL;
struct amdgpu_ctx *ctx;
struct fence *fence;
struct dma_fence *fence;
long r;
r = amdgpu_cs_get_ring(adev, wait->in.ip_type, wait->in.ip_instance,
......@@ -1107,8 +1124,8 @@ int amdgpu_cs_wait_ioctl(struct drm_device *dev, void *data,
if (IS_ERR(fence))
r = PTR_ERR(fence);
else if (fence) {
r = fence_wait_timeout(fence, true, timeout);
fence_put(fence);
r = dma_fence_wait_timeout(fence, true, timeout);
dma_fence_put(fence);
} else
r = 1;
......@@ -1195,6 +1212,15 @@ int amdgpu_cs_sysvm_access_required(struct amdgpu_cs_parser *parser)
r = amdgpu_ttm_bind(&bo->tbo, &bo->tbo.mem);
if (unlikely(r))
return r;
if (bo->flags & AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS)
continue;
bo->flags |= AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS;
amdgpu_ttm_placement_from_domain(bo, bo->allowed_domains);
r = ttm_bo_validate(&bo->tbo, &bo->placement, false, false);
if (unlikely(r))
return r;
}
return 0;
......
......@@ -35,7 +35,7 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev, struct amdgpu_ctx *ctx)
kref_init(&ctx->refcount);
spin_lock_init(&ctx->ring_lock);
ctx->fences = kcalloc(amdgpu_sched_jobs * AMDGPU_MAX_RINGS,
sizeof(struct fence*), GFP_KERNEL);
sizeof(struct dma_fence*), GFP_KERNEL);
if (!ctx->fences)
return -ENOMEM;
......@@ -55,18 +55,18 @@ static int amdgpu_ctx_init(struct amdgpu_device *adev, struct amdgpu_ctx *ctx)
r = amd_sched_entity_init(&ring->sched, &ctx->rings[i].entity,
rq, amdgpu_sched_jobs);
if (r)
break;
goto failed;
}
if (i < adev->num_rings) {
for (j = 0; j < i; j++)
amd_sched_entity_fini(&adev->rings[j]->sched,
&ctx->rings[j].entity);
kfree(ctx->fences);
ctx->fences = NULL;
return r;
}
return 0;
failed:
for (j = 0; j < i; j++)
amd_sched_entity_fini(&adev->rings[j]->sched,
&ctx->rings[j].entity);
kfree(ctx->fences);
ctx->fences = NULL;
return r;
}
static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
......@@ -79,7 +79,7 @@ static void amdgpu_ctx_fini(struct amdgpu_ctx *ctx)
for (i = 0; i < AMDGPU_MAX_RINGS; ++i)
for (j = 0; j < amdgpu_sched_jobs; ++j)
fence_put(ctx->rings[i].fences[j]);
dma_fence_put(ctx->rings[i].fences[j]);
kfree(ctx->fences);
ctx->fences = NULL;
......@@ -241,39 +241,39 @@ int amdgpu_ctx_put(struct amdgpu_ctx *ctx)
}
uint64_t amdgpu_ctx_add_fence(struct amdgpu_ctx *ctx, struct amdgpu_ring *ring,
struct fence *fence)
struct dma_fence *fence)
{
struct amdgpu_ctx_ring *cring = & ctx->rings[ring->idx];
uint64_t seq = cring->sequence;
unsigned idx = 0;
struct fence *other = NULL;
struct dma_fence *other = NULL;
idx = seq & (amdgpu_sched_jobs - 1);
other = cring->fences[idx];
if (other) {
signed long r;
r = fence_wait_timeout(other, false, MAX_SCHEDULE_TIMEOUT);
r = dma_fence_wait_timeout(other, false, MAX_SCHEDULE_TIMEOUT);
if (r < 0)
DRM_ERROR("Error (%ld) waiting for fence!\n", r);
}
fence_get(fence);
dma_fence_get(fence);
spin_lock(&ctx->ring_lock);
cring->fences[idx] = fence;
cring->sequence++;
spin_unlock(&ctx->ring_lock);
fence_put(other);
dma_fence_put(other);
return seq;
}
struct fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
struct amdgpu_ring *ring, uint64_t seq)
struct dma_fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
struct amdgpu_ring *ring, uint64_t seq)
{
struct amdgpu_ctx_ring *cring = & ctx->rings[ring->idx];
struct fence *fence;
struct dma_fence *fence;
spin_lock(&ctx->ring_lock);
......@@ -288,7 +288,7 @@ struct fence *amdgpu_ctx_get_fence(struct amdgpu_ctx *ctx,
return NULL;
}
fence = fence_get(cring->fences[seq & (amdgpu_sched_jobs - 1)]);
fence = dma_fence_get(cring->fences[seq & (amdgpu_sched_jobs - 1)]);
spin_unlock(&ctx->ring_lock);
return fence;
......
......@@ -35,29 +35,29 @@
#include <drm/drm_crtc_helper.h>
#include <drm/drm_edid.h>
static void amdgpu_flip_callback(struct fence *f, struct fence_cb *cb)
static void amdgpu_flip_callback(struct dma_fence *f, struct dma_fence_cb *cb)
{
struct amdgpu_flip_work *work =
container_of(cb, struct amdgpu_flip_work, cb);
fence_put(f);
dma_fence_put(f);
schedule_work(&work->flip_work.work);
}
static bool amdgpu_flip_handle_fence(struct amdgpu_flip_work *work,
struct fence **f)
struct dma_fence **f)
{
struct fence *fence= *f;
struct dma_fence *fence= *f;
if (fence == NULL)
return false;
*f = NULL;
if (!fence_add_callback(fence, &work->cb, amdgpu_flip_callback))
if (!dma_fence_add_callback(fence, &work->cb, amdgpu_flip_callback))
return true;
fence_put(fence);
dma_fence_put(fence);
return false;
}
......@@ -68,9 +68,9 @@ static void amdgpu_flip_work_func(struct work_struct *__work)
struct amdgpu_flip_work *work =
container_of(delayed_work, struct amdgpu_flip_work, flip_work);
struct amdgpu_device *adev = work->adev;
struct amdgpu_crtc *amdgpuCrtc = adev->mode_info.crtcs[work->crtc_id];
struct amdgpu_crtc *amdgpu_crtc = adev->mode_info.crtcs[work->crtc_id];
struct drm_crtc *crtc = &amdgpuCrtc->base;
struct drm_crtc *crtc = &amdgpu_crtc->base;
unsigned long flags;
unsigned i;
int vpos, hpos;
......@@ -85,14 +85,14 @@ static void amdgpu_flip_work_func(struct work_struct *__work)
/* Wait until we're out of the vertical blank period before the one
* targeted by the flip
*/
if (amdgpuCrtc->enabled &&
if (amdgpu_crtc->enabled &&
(amdgpu_get_crtc_scanoutpos(adev->ddev, work->crtc_id, 0,
&vpos, &hpos, NULL, NULL,
&crtc->hwmode)
& (DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK)) ==
(DRM_SCANOUTPOS_VALID | DRM_SCANOUTPOS_IN_VBLANK) &&
(int)(work->target_vblank -
amdgpu_get_vblank_counter_kms(adev->ddev, amdgpuCrtc->crtc_id)) > 0) {
amdgpu_get_vblank_counter_kms(adev->ddev, amdgpu_crtc->crtc_id)) > 0) {
schedule_delayed_work(&work->flip_work, usecs_to_jiffies(1000));
return;
}
......@@ -104,12 +104,12 @@ static void amdgpu_flip_work_func(struct work_struct *__work)
adev->mode_info.funcs->page_flip(adev, work->crtc_id, work->base, work->async);
/* Set the flip status */
amdgpuCrtc->pflip_status = AMDGPU_FLIP_SUBMITTED;
amdgpu_crtc->pflip_status = AMDGPU_FLIP_SUBMITTED;
spin_unlock_irqrestore(&crtc->dev->event_lock, flags);
DRM_DEBUG_DRIVER("crtc:%d[%p], pflip_stat:AMDGPU_FLIP_SUBMITTED, work: %p,\n",
amdgpuCrtc->crtc_id, amdgpuCrtc, work);
amdgpu_crtc->crtc_id, amdgpu_crtc, work);
}
......@@ -244,9 +244,9 @@ int amdgpu_crtc_page_flip_target(struct drm_crtc *crtc,
cleanup:
amdgpu_bo_unref(&work->old_abo);
fence_put(work->excl);
dma_fence_put(work->excl);
for (i = 0; i < work->shared_count; ++i)
fence_put(work->shared[i]);
dma_fence_put(work->shared[i]);
kfree(work->shared);
kfree(work);
......
......@@ -553,9 +553,10 @@ int amdgpu_parse_extended_power_table(struct amdgpu_device *adev)
entry = (ATOM_PPLIB_VCE_Clock_Voltage_Limit_Record *)
((u8 *)entry + sizeof(ATOM_PPLIB_VCE_Clock_Voltage_Limit_Record));
}
for (i = 0; i < states->numEntries; i++) {
if (i >= AMDGPU_MAX_VCE_LEVELS)
break;
adev->pm.dpm.num_of_vce_states =
states->numEntries > AMD_MAX_VCE_LEVELS ?
AMD_MAX_VCE_LEVELS : states->numEntries;
for (i = 0; i < adev->pm.dpm.num_of_vce_states; i++) {
vce_clk = (VCEClockInfo *)
((u8 *)&array->entries[0] +
(state_entry->ucVCEClockInfoIndex * sizeof(VCEClockInfo)));
......@@ -955,3 +956,12 @@ u8 amdgpu_encode_pci_lane_width(u32 lanes)
return encoded_lanes[lanes];
}
struct amd_vce_state*
amdgpu_get_vce_clock_state(struct amdgpu_device *adev, unsigned idx)
{
if (idx < adev->pm.dpm.num_of_vce_states)
return &adev->pm.dpm.vce_states[idx];
return NULL;
}
......@@ -23,6 +23,446 @@
#ifndef __AMDGPU_DPM_H__
#define __AMDGPU_DPM_H__
enum amdgpu_int_thermal_type {
THERMAL_TYPE_NONE,
THERMAL_TYPE_EXTERNAL,
THERMAL_TYPE_EXTERNAL_GPIO,
THERMAL_TYPE_RV6XX,
THERMAL_TYPE_RV770,
THERMAL_TYPE_ADT7473_WITH_INTERNAL,
THERMAL_TYPE_EVERGREEN,
THERMAL_TYPE_SUMO,
THERMAL_TYPE_NI,
THERMAL_TYPE_SI,
THERMAL_TYPE_EMC2103_WITH_INTERNAL,
THERMAL_TYPE_CI,
THERMAL_TYPE_KV,
};
enum amdgpu_dpm_auto_throttle_src {
AMDGPU_DPM_AUTO_THROTTLE_SRC_THERMAL,
AMDGPU_DPM_AUTO_THROTTLE_SRC_EXTERNAL
};
enum amdgpu_dpm_event_src {
AMDGPU_DPM_EVENT_SRC_ANALOG = 0,
AMDGPU_DPM_EVENT_SRC_EXTERNAL = 1,
AMDGPU_DPM_EVENT_SRC_DIGITAL = 2,
AMDGPU_DPM_EVENT_SRC_ANALOG_OR_EXTERNAL = 3,
AMDGPU_DPM_EVENT_SRC_DIGIAL_OR_EXTERNAL = 4
};
struct amdgpu_ps {
u32 caps; /* vbios flags */
u32 class; /* vbios flags */
u32 class2; /* vbios flags */
/* UVD clocks */
u32 vclk;
u32 dclk;
/* VCE clocks */
u32 evclk;
u32 ecclk;
bool vce_active;
enum amd_vce_level vce_level;
/* asic priv */
void *ps_priv;
};
struct amdgpu_dpm_thermal {
/* thermal interrupt work */
struct work_struct work;
/* low temperature threshold */
int min_temp;
/* high temperature threshold */
int max_temp;
/* was last interrupt low to high or high to low */
bool high_to_low;
/* interrupt source */
struct amdgpu_irq_src irq;
};
enum amdgpu_clk_action
{
AMDGPU_SCLK_UP = 1,
AMDGPU_SCLK_DOWN
};
struct amdgpu_blacklist_clocks
{
u32 sclk;
u32 mclk;
enum amdgpu_clk_action action;
};
struct amdgpu_clock_and_voltage_limits {
u32 sclk;
u32 mclk;
u16 vddc;
u16 vddci;
};
struct amdgpu_clock_array {
u32 count;
u32 *values;
};
struct amdgpu_clock_voltage_dependency_entry {
u32 clk;
u16 v;
};
struct amdgpu_clock_voltage_dependency_table {
u32 count;
struct amdgpu_clock_voltage_dependency_entry *entries;
};
union amdgpu_cac_leakage_entry {
struct {
u16 vddc;
u32 leakage;
};
struct {
u16 vddc1;
u16 vddc2;
u16 vddc3;
};
};
struct amdgpu_cac_leakage_table {
u32 count;
union amdgpu_cac_leakage_entry *entries;
};
struct amdgpu_phase_shedding_limits_entry {
u16 voltage;
u32 sclk;
u32 mclk;
};
struct amdgpu_phase_shedding_limits_table {
u32 count;
struct amdgpu_phase_shedding_limits_entry *entries;
};
struct amdgpu_uvd_clock_voltage_dependency_entry {
u32 vclk;
u32 dclk;
u16 v;
};
struct amdgpu_uvd_clock_voltage_dependency_table {
u8 count;
struct amdgpu_uvd_clock_voltage_dependency_entry *entries;
};
struct amdgpu_vce_clock_voltage_dependency_entry {
u32 ecclk;
u32 evclk;
u16 v;
};
struct amdgpu_vce_clock_voltage_dependency_table {
u8 count;
struct amdgpu_vce_clock_voltage_dependency_entry *entries;
};
struct amdgpu_ppm_table {
u8 ppm_design;
u16 cpu_core_number;
u32 platform_tdp;
u32 small_ac_platform_tdp;
u32 platform_tdc;
u32 small_ac_platform_tdc;
u32 apu_tdp;
u32 dgpu_tdp;
u32 dgpu_ulv_power;
u32 tj_max;
};
struct amdgpu_cac_tdp_table {
u16 tdp;
u16 configurable_tdp;
u16 tdc;
u16 battery_power_limit;
u16 small_power_limit;
u16 low_cac_leakage;
u16 high_cac_leakage;
u16 maximum_power_delivery_limit;
};
struct amdgpu_dpm_dynamic_state {
struct amdgpu_clock_voltage_dependency_table vddc_dependency_on_sclk;
struct amdgpu_clock_voltage_dependency_table vddci_dependency_on_mclk;
struct amdgpu_clock_voltage_dependency_table vddc_dependency_on_mclk;
struct amdgpu_clock_voltage_dependency_table mvdd_dependency_on_mclk;
struct amdgpu_clock_voltage_dependency_table vddc_dependency_on_dispclk;
struct amdgpu_uvd_clock_voltage_dependency_table uvd_clock_voltage_dependency_table;
struct amdgpu_vce_clock_voltage_dependency_table vce_clock_voltage_dependency_table;
struct amdgpu_clock_voltage_dependency_table samu_clock_voltage_dependency_table;
struct amdgpu_clock_voltage_dependency_table acp_clock_voltage_dependency_table;
struct amdgpu_clock_voltage_dependency_table vddgfx_dependency_on_sclk;
struct amdgpu_clock_array valid_sclk_values;
struct amdgpu_clock_array valid_mclk_values;
struct amdgpu_clock_and_voltage_limits max_clock_voltage_on_dc;
struct amdgpu_clock_and_voltage_limits max_clock_voltage_on_ac;
u32 mclk_sclk_ratio;
u32 sclk_mclk_delta;
u16 vddc_vddci_delta;
u16 min_vddc_for_pcie_gen2;
struct amdgpu_cac_leakage_table cac_leakage_table;
struct amdgpu_phase_shedding_limits_table phase_shedding_limits_table;
struct amdgpu_ppm_table *ppm_table;
struct amdgpu_cac_tdp_table *cac_tdp_table;
};
struct amdgpu_dpm_fan {
u16 t_min;
u16 t_med;
u16 t_high;
u16 pwm_min;
u16 pwm_med;
u16 pwm_high;
u8 t_hyst;
u32 cycle_delay;
u16 t_max;
u8 control_mode;
u16 default_max_fan_pwm;
u16 default_fan_output_sensitivity;
u16 fan_output_sensitivity;
bool ucode_fan_control;
};
enum amdgpu_pcie_gen {
AMDGPU_PCIE_GEN1 = 0,
AMDGPU_PCIE_GEN2 = 1,
AMDGPU_PCIE_GEN3 = 2,
AMDGPU_PCIE_GEN_INVALID = 0xffff
};
enum amdgpu_dpm_forced_level {
AMDGPU_DPM_FORCED_LEVEL_AUTO = 0,
AMDGPU_DPM_FORCED_LEVEL_LOW = 1,
AMDGPU_DPM_FORCED_LEVEL_HIGH = 2,
AMDGPU_DPM_FORCED_LEVEL_MANUAL = 3,
};
struct amdgpu_dpm_funcs {
int (*get_temperature)(struct amdgpu_device *adev);
int (*pre_set_power_state)(struct amdgpu_device *adev);
int (*set_power_state)(struct amdgpu_device *adev);
void (*post_set_power_state)(struct amdgpu_device *adev);
void (*display_configuration_changed)(struct amdgpu_device *adev);
u32 (*get_sclk)(struct amdgpu_device *adev, bool low);
u32 (*get_mclk)(struct amdgpu_device *adev, bool low);
void (*print_power_state)(struct amdgpu_device *adev, struct amdgpu_ps *ps);
void (*debugfs_print_current_performance_level)(struct amdgpu_device *adev, struct seq_file *m);
int (*force_performance_level)(struct amdgpu_device *adev, enum amdgpu_dpm_forced_level level);
bool (*vblank_too_short)(struct amdgpu_device *adev);
void (*powergate_uvd)(struct amdgpu_device *adev, bool gate);
void (*powergate_vce)(struct amdgpu_device *adev, bool gate);
void (*enable_bapm)(struct amdgpu_device *adev, bool enable);
void (*set_fan_control_mode)(struct amdgpu_device *adev, u32 mode);
u32 (*get_fan_control_mode)(struct amdgpu_device *adev);
int (*set_fan_speed_percent)(struct amdgpu_device *adev, u32 speed);
int (*get_fan_speed_percent)(struct amdgpu_device *adev, u32 *speed);
int (*force_clock_level)(struct amdgpu_device *adev, enum pp_clock_type type, uint32_t mask);
int (*print_clock_levels)(struct amdgpu_device *adev, enum pp_clock_type type, char *buf);
int (*get_sclk_od)(struct amdgpu_device *adev);
int (*set_sclk_od)(struct amdgpu_device *adev, uint32_t value);
int (*get_mclk_od)(struct amdgpu_device *adev);
int (*set_mclk_od)(struct amdgpu_device *adev, uint32_t value);
int (*check_state_equal)(struct amdgpu_device *adev,
struct amdgpu_ps *cps,
struct amdgpu_ps *rps,
bool *equal);
struct amd_vce_state* (*get_vce_clock_state)(struct amdgpu_device *adev, unsigned idx);
};
#define amdgpu_dpm_pre_set_power_state(adev) (adev)->pm.funcs->pre_set_power_state((adev))
#define amdgpu_dpm_set_power_state(adev) (adev)->pm.funcs->set_power_state((adev))
#define amdgpu_dpm_post_set_power_state(adev) (adev)->pm.funcs->post_set_power_state((adev))
#define amdgpu_dpm_display_configuration_changed(adev) (adev)->pm.funcs->display_configuration_changed((adev))
#define amdgpu_dpm_print_power_state(adev, ps) (adev)->pm.funcs->print_power_state((adev), (ps))
#define amdgpu_dpm_vblank_too_short(adev) (adev)->pm.funcs->vblank_too_short((adev))
#define amdgpu_dpm_enable_bapm(adev, e) (adev)->pm.funcs->enable_bapm((adev), (e))
#define amdgpu_dpm_read_sensor(adev, idx, value) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->read_sensor(adev->powerplay.pp_handle, (idx), (value)) : \
-EINVAL)
#define amdgpu_dpm_get_temperature(adev) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->get_temperature((adev)->powerplay.pp_handle) : \
(adev)->pm.funcs->get_temperature((adev)))
#define amdgpu_dpm_set_fan_control_mode(adev, m) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->set_fan_control_mode((adev)->powerplay.pp_handle, (m)) : \
(adev)->pm.funcs->set_fan_control_mode((adev), (m)))
#define amdgpu_dpm_get_fan_control_mode(adev) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->get_fan_control_mode((adev)->powerplay.pp_handle) : \
(adev)->pm.funcs->get_fan_control_mode((adev)))
#define amdgpu_dpm_set_fan_speed_percent(adev, s) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->set_fan_speed_percent((adev)->powerplay.pp_handle, (s)) : \
(adev)->pm.funcs->set_fan_speed_percent((adev), (s)))
#define amdgpu_dpm_get_fan_speed_percent(adev, s) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->get_fan_speed_percent((adev)->powerplay.pp_handle, (s)) : \
(adev)->pm.funcs->get_fan_speed_percent((adev), (s)))
#define amdgpu_dpm_get_sclk(adev, l) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->get_sclk((adev)->powerplay.pp_handle, (l)) : \
(adev)->pm.funcs->get_sclk((adev), (l)))
#define amdgpu_dpm_get_mclk(adev, l) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->get_mclk((adev)->powerplay.pp_handle, (l)) : \
(adev)->pm.funcs->get_mclk((adev), (l)))
#define amdgpu_dpm_force_performance_level(adev, l) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->force_performance_level((adev)->powerplay.pp_handle, (l)) : \
(adev)->pm.funcs->force_performance_level((adev), (l)))
#define amdgpu_dpm_powergate_uvd(adev, g) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->powergate_uvd((adev)->powerplay.pp_handle, (g)) : \
(adev)->pm.funcs->powergate_uvd((adev), (g)))
#define amdgpu_dpm_powergate_vce(adev, g) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->powergate_vce((adev)->powerplay.pp_handle, (g)) : \
(adev)->pm.funcs->powergate_vce((adev), (g)))
#define amdgpu_dpm_get_current_power_state(adev) \
(adev)->powerplay.pp_funcs->get_current_power_state((adev)->powerplay.pp_handle)
#define amdgpu_dpm_get_performance_level(adev) \
(adev)->powerplay.pp_funcs->get_performance_level((adev)->powerplay.pp_handle)
#define amdgpu_dpm_get_pp_num_states(adev, data) \
(adev)->powerplay.pp_funcs->get_pp_num_states((adev)->powerplay.pp_handle, data)
#define amdgpu_dpm_get_pp_table(adev, table) \
(adev)->powerplay.pp_funcs->get_pp_table((adev)->powerplay.pp_handle, table)
#define amdgpu_dpm_set_pp_table(adev, buf, size) \
(adev)->powerplay.pp_funcs->set_pp_table((adev)->powerplay.pp_handle, buf, size)
#define amdgpu_dpm_print_clock_levels(adev, type, buf) \
(adev)->powerplay.pp_funcs->print_clock_levels((adev)->powerplay.pp_handle, type, buf)
#define amdgpu_dpm_force_clock_level(adev, type, level) \
(adev)->powerplay.pp_funcs->force_clock_level((adev)->powerplay.pp_handle, type, level)
#define amdgpu_dpm_get_sclk_od(adev) \
(adev)->powerplay.pp_funcs->get_sclk_od((adev)->powerplay.pp_handle)
#define amdgpu_dpm_set_sclk_od(adev, value) \
(adev)->powerplay.pp_funcs->set_sclk_od((adev)->powerplay.pp_handle, value)
#define amdgpu_dpm_get_mclk_od(adev) \
((adev)->powerplay.pp_funcs->get_mclk_od((adev)->powerplay.pp_handle))
#define amdgpu_dpm_set_mclk_od(adev, value) \
((adev)->powerplay.pp_funcs->set_mclk_od((adev)->powerplay.pp_handle, value))
#define amdgpu_dpm_dispatch_task(adev, event_id, input, output) \
(adev)->powerplay.pp_funcs->dispatch_tasks((adev)->powerplay.pp_handle, (event_id), (input), (output))
#define amgdpu_dpm_check_state_equal(adev, cps, rps, equal) (adev)->pm.funcs->check_state_equal((adev), (cps),(rps),(equal))
#define amdgpu_dpm_get_vce_clock_state(adev, i) \
((adev)->pp_enabled ? \
(adev)->powerplay.pp_funcs->get_vce_clock_state((adev)->powerplay.pp_handle, (i)) : \
(adev)->pm.funcs->get_vce_clock_state((adev), (i)))
struct amdgpu_dpm {
struct amdgpu_ps *ps;
/* number of valid power states */
int num_ps;
/* current power state that is active */
struct amdgpu_ps *current_ps;
/* requested power state */
struct amdgpu_ps *requested_ps;
/* boot up power state */
struct amdgpu_ps *boot_ps;
/* default uvd power state */
struct amdgpu_ps *uvd_ps;
/* vce requirements */
u32 num_of_vce_states;
struct amd_vce_state vce_states[AMD_MAX_VCE_LEVELS];
enum amd_vce_level vce_level;
enum amd_pm_state_type state;
enum amd_pm_state_type user_state;
enum amd_pm_state_type last_state;
enum amd_pm_state_type last_user_state;
u32 platform_caps;
u32 voltage_response_time;
u32 backbias_response_time;
void *priv;
u32 new_active_crtcs;
int new_active_crtc_count;
u32 current_active_crtcs;
int current_active_crtc_count;
struct amdgpu_dpm_dynamic_state dyn_state;
struct amdgpu_dpm_fan fan;
u32 tdp_limit;
u32 near_tdp_limit;
u32 near_tdp_limit_adjusted;
u32 sq_ramping_threshold;
u32 cac_leakage;
u16 tdp_od_limit;
u32 tdp_adjustment;
u16 load_line_slope;
bool power_control;
bool ac_power;
/* special states active */
bool thermal_active;
bool uvd_active;
bool vce_active;
/* thermal handling */
struct amdgpu_dpm_thermal thermal;
/* forced levels */
enum amdgpu_dpm_forced_level forced_level;
};
struct amdgpu_pm {
struct mutex mutex;
u32 current_sclk;
u32 current_mclk;
u32 default_sclk;
u32 default_mclk;
struct amdgpu_i2c_chan *i2c_bus;
/* internal thermal controller on rv6xx+ */
enum amdgpu_int_thermal_type int_thermal_type;
struct device *int_hwmon_dev;
/* fan control parameters */
bool no_fan;
u8 fan_pulses_per_revolution;
u8 fan_min_rpm;
u8 fan_max_rpm;
/* dpm */
bool dpm_enabled;
bool sysfs_initialized;
struct amdgpu_dpm dpm;
const struct firmware *fw; /* SMC firmware */
uint32_t fw_version;
const struct amdgpu_dpm_funcs *funcs;
uint32_t pcie_gen_mask;
uint32_t pcie_mlw_mask;
struct amd_pp_display_configuration pm_display_cfg;/* set by DAL */
};
#define R600_SSTU_DFLT 0
#define R600_SST_DFLT 0x00C8
......@@ -82,4 +522,7 @@ u16 amdgpu_get_pcie_lane_support(struct amdgpu_device *adev,
u16 default_lanes);
u8 amdgpu_encode_pci_lane_width(u32 lanes);
struct amd_vce_state*
amdgpu_get_vce_clock_state(struct amdgpu_device *adev, unsigned idx);
#endif
......@@ -58,9 +58,10 @@
* - 3.6.0 - kmd involves use CONTEXT_CONTROL in ring buffer.
* - 3.7.0 - Add support for VCE clock list packet
* - 3.8.0 - Add support raster config init in the kernel
* - 3.9.0 - Add support for memory query info about VRAM and GTT.
*/
#define KMS_DRIVER_MAJOR 3
#define KMS_DRIVER_MINOR 8
#define KMS_DRIVER_MINOR 9
#define KMS_DRIVER_PATCHLEVEL 0
int amdgpu_vram_limit = 0;
......@@ -85,6 +86,7 @@ int amdgpu_vm_size = 64;
int amdgpu_vm_block_size = -1;
int amdgpu_vm_fault_stop = 0;
int amdgpu_vm_debug = 0;
int amdgpu_vram_page_split = 1024;
int amdgpu_exp_hw_support = 0;
int amdgpu_sched_jobs = 32;
int amdgpu_sched_hw_submission = 2;
......@@ -165,6 +167,9 @@ module_param_named(vm_fault_stop, amdgpu_vm_fault_stop, int, 0444);
MODULE_PARM_DESC(vm_debug, "Debug VM handling (0 = disabled (default), 1 = enabled)");
module_param_named(vm_debug, amdgpu_vm_debug, int, 0644);
MODULE_PARM_DESC(vram_page_split, "Number of pages after we split VRAM allocations (default 1024, -1 = disable)");
module_param_named(vram_page_split, amdgpu_vram_page_split, int, 0444);
MODULE_PARM_DESC(exp_hw_support, "experimental hw support (1 = enable, 0 = disable (default))");
module_param_named(exp_hw_support, amdgpu_exp_hw_support, int, 0444);
......@@ -201,7 +206,8 @@ module_param_named(pg_mask, amdgpu_pg_mask, uint, 0444);
MODULE_PARM_DESC(disable_cu, "Disable CUs (se.sh.cu,...)");
module_param_named(disable_cu, amdgpu_disable_cu, charp, 0444);
MODULE_PARM_DESC(virtual_display, "Enable virtual display feature (the virtual_display will be set like xxxx:xx:xx.x;xxxx:xx:xx.x)");
MODULE_PARM_DESC(virtual_display,
"Enable virtual display feature (the virtual_display will be set like xxxx:xx:xx.x,x;xxxx:xx:xx.x,x)");
module_param_named(virtual_display, amdgpu_virtual_display, charp, 0444);
static const struct pci_device_id pciidlist[] = {
......@@ -381,6 +387,7 @@ static const struct pci_device_id pciidlist[] = {
{0x1002, 0x6939, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_TONGA},
/* fiji */
{0x1002, 0x7300, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_FIJI},
{0x1002, 0x730F, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_FIJI},
/* carrizo */
{0x1002, 0x9870, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_CARRIZO|AMD_IS_APU},
{0x1002, 0x9874, PCI_ANY_ID, PCI_ANY_ID, 0, 0, CHIP_CARRIZO|AMD_IS_APU},
......
......@@ -152,7 +152,8 @@ static int amdgpufb_create_pinned_object(struct amdgpu_fbdev *rfbdev,
aligned_size = ALIGN(size, PAGE_SIZE);
ret = amdgpu_gem_object_create(adev, aligned_size, 0,
AMDGPU_GEM_DOMAIN_VRAM,
AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
true, &gobj);
if (ret) {
printk(KERN_ERR "failed to allocate framebuffer (%d)\n",
......
......@@ -126,7 +126,8 @@ int amdgpu_gart_table_vram_alloc(struct amdgpu_device *adev)
if (adev->gart.robj == NULL) {
r = amdgpu_bo_create(adev, adev->gart.table_size,
PAGE_SIZE, true, AMDGPU_GEM_DOMAIN_VRAM,
AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED,
AMDGPU_GEM_CREATE_CPU_ACCESS_REQUIRED |
AMDGPU_GEM_CREATE_VRAM_CONTIGUOUS,
NULL, NULL, &adev->gart.robj);
if (r) {
return r;
......
......@@ -116,10 +116,11 @@ void amdgpu_gem_force_release(struct amdgpu_device *adev)
* Call from drm_gem_handle_create which appear in both new and open ioctl
* case.
*/
int amdgpu_gem_object_open(struct drm_gem_object *obj, struct drm_file *file_priv)
int amdgpu_gem_object_open(struct drm_gem_object *obj,
struct drm_file *file_priv)
{
struct amdgpu_bo *abo = gem_to_amdgpu_bo(obj);
struct amdgpu_device *adev = abo->adev;
struct amdgpu_device *adev = amdgpu_ttm_adev(abo->tbo.bdev);
struct amdgpu_fpriv *fpriv = file_priv->driver_priv;
struct amdgpu_vm *vm = &fpriv->vm;
struct amdgpu_bo_va *bo_va;
......@@ -142,7 +143,7 @@ void amdgpu_gem_object_close(struct drm_gem_object *obj,
struct drm_file *file_priv)
{
struct amdgpu_bo *bo = gem_to_amdgpu_bo(obj);
struct amdgpu_device *adev = bo->adev;
struct amdgpu_device *adev = amdgpu_ttm_adev(bo->tbo.bdev);
struct amdgpu_fpriv *fpriv = file_priv->driver_priv;
struct amdgpu_vm *vm = &fpriv->vm;
......@@ -468,6 +469,16 @@ int amdgpu_gem_metadata_ioctl(struct drm_device *dev, void *data,
return r;
}
static int amdgpu_gem_va_check(void *param, struct amdgpu_bo *bo)
{
unsigned domain = amdgpu_mem_type_to_domain(bo->tbo.mem.mem_type);
/* if anything is swapped out don't swap it in here,
just abort and wait for the next CS */
return domain == AMDGPU_GEM_DOMAIN_CPU ? -ERESTARTSYS : 0;
}
/**
* amdgpu_gem_va_update_vm -update the bo_va in its VM
*
......@@ -478,7 +489,8 @@ int amdgpu_gem_metadata_ioctl(struct drm_device *dev, void *data,
* vital here, so they are not reported back to userspace.
*/
static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
struct amdgpu_bo_va *bo_va, uint32_t operation)
struct amdgpu_bo_va *bo_va,
uint32_t operation)
{
struct ttm_validate_buffer tv, *entry;
struct amdgpu_bo_list_entry vm_pd;
......@@ -501,7 +513,6 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
if (r)
goto error_print;
amdgpu_vm_get_pt_bos(adev, bo_va->vm, &duplicates);
list_for_each_entry(entry, &list, head) {
domain = amdgpu_mem_type_to_domain(entry->bo->mem.mem_type);
/* if anything is swapped out don't swap it in here,
......@@ -509,13 +520,10 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
if (domain == AMDGPU_GEM_DOMAIN_CPU)
goto error_unreserve;
}
list_for_each_entry(entry, &duplicates, head) {
domain = amdgpu_mem_type_to_domain(entry->bo->mem.mem_type);
/* if anything is swapped out don't swap it in here,
just abort and wait for the next CS */
if (domain == AMDGPU_GEM_DOMAIN_CPU)
goto error_unreserve;
}
r = amdgpu_vm_validate_pt_bos(adev, bo_va->vm, amdgpu_gem_va_check,
NULL);
if (r)
goto error_unreserve;
r = amdgpu_vm_update_page_directory(adev, bo_va->vm);
if (r)
......@@ -536,8 +544,6 @@ static void amdgpu_gem_va_update_vm(struct amdgpu_device *adev,
DRM_ERROR("Couldn't update BO_VA (%d)\n", r);
}
int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
struct drm_file *filp)
{
......@@ -547,7 +553,8 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
struct amdgpu_fpriv *fpriv = filp->driver_priv;
struct amdgpu_bo *abo;
struct amdgpu_bo_va *bo_va;
struct ttm_validate_buffer tv, tv_pd;
struct amdgpu_bo_list_entry vm_pd;
struct ttm_validate_buffer tv;
struct ww_acquire_ctx ticket;
struct list_head list, duplicates;
uint32_t invalid_flags, va_flags = 0;
......@@ -592,9 +599,7 @@ int amdgpu_gem_va_ioctl(struct drm_device *dev, void *data,
tv.shared = true;
list_add(&tv.head, &list);
tv_pd.bo = &fpriv->vm.page_directory->tbo;
tv_pd.shared = true;
list_add(&tv_pd.head, &list);
amdgpu_vm_get_pd_bo(&fpriv->vm, &list, &vm_pd);
r = ttm_eu_reserve_buffers(&ticket, &list, true, &duplicates);
if (r) {
......
......@@ -24,6 +24,7 @@
*/
#include <drm/drmP.h>
#include "amdgpu.h"
#include "amdgpu_gfx.h"
/*
* GPU scratch registers helpers function.
......
......@@ -27,6 +27,7 @@
int amdgpu_gfx_scratch_get(struct amdgpu_device *adev, uint32_t *reg);
void amdgpu_gfx_scratch_free(struct amdgpu_device *adev, uint32_t reg);
unsigned amdgpu_gfx_parse_disable_cu(unsigned *mask, unsigned max_se, unsigned max_sh);
void amdgpu_gfx_parse_disable_cu(unsigned *mask, unsigned max_se,
unsigned max_sh);
#endif
......@@ -168,6 +168,7 @@ static int amdgpu_gtt_mgr_new(struct ttm_mem_type_manager *man,
return -ENOMEM;
node->start = AMDGPU_BO_INVALID_OFFSET;
node->size = mem->num_pages;
mem->mm_node = node;
if (place->fpfn || place->lpfn || place->flags & TTM_PL_FLAG_TOPDOWN) {
......
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
此差异已折叠。
Markdown is supported
0% .
You are about to add 0 people to the discussion. Proceed with caution.
先完成此消息的编辑!
想要评论请 注册