linux_dsm_epyc7002/drivers/char/drm/drm_fops.c

470 lines
12 KiB
C
Raw Normal View History

/**
* \file drm_fops.c
* File operations for DRM
*
* \author Rickard E. (Rik) Faith <faith@valinux.com>
* \author Daryll Strauss <daryll@valinux.com>
* \author Gareth Hughes <gareth@valinux.com>
*/
/*
* Created: Mon Jan 4 08:58:31 1999 by faith@valinux.com
*
* Copyright 1999 Precision Insight, Inc., Cedar Park, Texas.
* Copyright 2000 VA Linux Systems, Inc., Sunnyvale, California.
* All Rights Reserved.
*
* Permission is hereby granted, free of charge, to any person obtaining a
* copy of this software and associated documentation files (the "Software"),
* to deal in the Software without restriction, including without limitation
* the rights to use, copy, modify, merge, publish, distribute, sublicense,
* and/or sell copies of the Software, and to permit persons to whom the
* Software is furnished to do so, subject to the following conditions:
*
* The above copyright notice and this permission notice (including the next
* paragraph) shall be included in all copies or substantial portions of the
* Software.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL
* VA LINUX SYSTEMS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM, DAMAGES OR
* OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
* ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
* OTHER DEALINGS IN THE SOFTWARE.
*/
#include "drmP.h"
#include "drm_sarea.h"
#include <linux/poll.h>
static int drm_open_helper(struct inode *inode, struct file *filp,
struct drm_device * dev);
static int drm_setup(struct drm_device * dev)
{
drm_local_map_t *map;
int i;
int ret;
u32 sareapage;
if (dev->driver->firstopen) {
ret = dev->driver->firstopen(dev);
if (ret != 0)
return ret;
}
dev->magicfree.next = NULL;
/* prebuild the SAREA */
sareapage = max_t(unsigned, SAREA_MAX, PAGE_SIZE);
i = drm_addmap(dev, 0, sareapage, _DRM_SHM, _DRM_CONTAINS_LOCK, &map);
if (i != 0)
return i;
atomic_set(&dev->ioctl_count, 0);
atomic_set(&dev->vma_count, 0);
dev->buf_use = 0;
atomic_set(&dev->buf_alloc, 0);
if (drm_core_check_feature(dev, DRIVER_HAVE_DMA)) {
i = drm_dma_setup(dev);
if (i < 0)
return i;
}
for (i = 0; i < ARRAY_SIZE(dev->counts); i++)
atomic_set(&dev->counts[i], 0);
drm_ht_create(&dev->magiclist, DRM_MAGIC_HASH_ORDER);
INIT_LIST_HEAD(&dev->magicfree);
dev->sigdata.lock = NULL;
init_waitqueue_head(&dev->lock.lock_queue);
dev->queue_count = 0;
dev->queue_reserved = 0;
dev->queue_slots = 0;
dev->queuelist = NULL;
dev->irq_enabled = 0;
dev->context_flag = 0;
dev->interrupt_flag = 0;
dev->dma_flag = 0;
dev->last_context = 0;
dev->last_switch = 0;
dev->last_checked = 0;
init_waitqueue_head(&dev->context_wait);
dev->if_version = 0;
dev->ctx_start = 0;
dev->lck_start = 0;
dev->buf_async = NULL;
init_waitqueue_head(&dev->buf_readers);
init_waitqueue_head(&dev->buf_writers);
DRM_DEBUG("\n");
/*
* The kernel's context could be created here, but is now created
* in drm_dma_enqueue. This is more resource-efficient for
* hardware that does not do DMA, but may mean that
* drm_select_queue fails between the time the interrupt is
* initialized and the time the queues are initialized.
*/
return 0;
}
/**
* Open file.
*
* \param inode device inode
* \param filp file pointer.
* \return zero on success or a negative number on failure.
*
* Searches the DRM device with the same minor number, calls open_helper(), and
* increments the device open count. If the open count was previous at zero,
* i.e., it's the first that the device is open, then calls setup().
*/
int drm_open(struct inode *inode, struct file *filp)
{
struct drm_device *dev = NULL;
int minor_id = iminor(inode);
struct drm_minor *minor;
int retcode = 0;
minor = idr_find(&drm_minors_idr, minor_id);
if (!minor)
return -ENODEV;
if (!(dev = minor->dev))
return -ENODEV;
retcode = drm_open_helper(inode, filp, dev);
if (!retcode) {
atomic_inc(&dev->counts[_DRM_STAT_OPENS]);
spin_lock(&dev->count_lock);
if (!dev->open_count++) {
spin_unlock(&dev->count_lock);
return drm_setup(dev);
}
spin_unlock(&dev->count_lock);
}
return retcode;
}
EXPORT_SYMBOL(drm_open);
/**
* File \c open operation.
*
* \param inode device inode.
* \param filp file pointer.
*
* Puts the dev->fops corresponding to the device minor number into
* \p filp, call the \c open method, and restore the file operations.
*/
int drm_stub_open(struct inode *inode, struct file *filp)
{
struct drm_device *dev = NULL;
struct drm_minor *minor;
int minor_id = iminor(inode);
int err = -ENODEV;
const struct file_operations *old_fops;
DRM_DEBUG("\n");
minor = idr_find(&drm_minors_idr, minor_id);
if (!minor)
return -ENODEV;
if (!(dev = minor->dev))
return -ENODEV;
old_fops = filp->f_op;
filp->f_op = fops_get(&dev->driver->fops);
if (filp->f_op->open && (err = filp->f_op->open(inode, filp))) {
fops_put(filp->f_op);
filp->f_op = fops_get(old_fops);
}
fops_put(old_fops);
return err;
}
/**
* Check whether DRI will run on this CPU.
*
* \return non-zero if the DRI will run on this CPU, or zero otherwise.
*/
static int drm_cpu_valid(void)
{
#if defined(__i386__)
if (boot_cpu_data.x86 == 3)
return 0; /* No cmpxchg on a 386 */
#endif
#if defined(__sparc__) && !defined(__sparc_v9__)
return 0; /* No cmpxchg before v9 sparc. */
#endif
return 1;
}
/**
* Called whenever a process opens /dev/drm.
*
* \param inode device inode.
* \param filp file pointer.
* \param dev device.
* \return zero on success or a negative number on failure.
*
* Creates and initializes a drm_file structure for the file private data in \p
* filp and add it into the double linked list in \p dev.
*/
static int drm_open_helper(struct inode *inode, struct file *filp,
struct drm_device * dev)
{
int minor_id = iminor(inode);
struct drm_file *priv;
int ret;
if (filp->f_flags & O_EXCL)
return -EBUSY; /* No exclusive opens */
if (!drm_cpu_valid())
return -EINVAL;
DRM_DEBUG("pid = %d, minor = %d\n", task_pid_nr(current), minor_id);
priv = drm_alloc(sizeof(*priv), DRM_MEM_FILES);
if (!priv)
return -ENOMEM;
memset(priv, 0, sizeof(*priv));
filp->private_data = priv;
priv->filp = filp;
priv->uid = current->euid;
priv->pid = task_pid_nr(current);
priv->minor = idr_find(&drm_minors_idr, minor_id);
priv->ioctl_count = 0;
/* for compatibility root is always authenticated */
priv->authenticated = capable(CAP_SYS_ADMIN);
priv->lock_count = 0;
INIT_LIST_HEAD(&priv->lhead);
if (dev->driver->open) {
ret = dev->driver->open(dev, priv);
if (ret < 0)
goto out_free;
}
mutex_lock(&dev->struct_mutex);
if (list_empty(&dev->filelist))
priv->master = 1;
list_add(&priv->lhead, &dev->filelist);
mutex_unlock(&dev->struct_mutex);
#ifdef __alpha__
/*
* Default the hose
*/
if (!dev->hose) {
struct pci_dev *pci_dev;
pci_dev = pci_get_class(PCI_CLASS_DISPLAY_VGA << 8, NULL);
if (pci_dev) {
dev->hose = pci_dev->sysdata;
pci_dev_put(pci_dev);
}
if (!dev->hose) {
struct pci_bus *b = pci_bus_b(pci_root_buses.next);
if (b)
dev->hose = b->sysdata;
}
}
#endif
return 0;
out_free:
drm_free(priv, sizeof(*priv), DRM_MEM_FILES);
filp->private_data = NULL;
return ret;
}
/** No-op. */
int drm_fasync(int fd, struct file *filp, int on)
{
struct drm_file *priv = filp->private_data;
struct drm_device *dev = priv->minor->dev;
int retcode;
DRM_DEBUG("fd = %d, device = 0x%lx\n", fd,
(long)old_encode_dev(priv->minor->device));
retcode = fasync_helper(fd, filp, on, &dev->buf_async);
if (retcode < 0)
return retcode;
return 0;
}
EXPORT_SYMBOL(drm_fasync);
/**
* Release file.
*
* \param inode device inode
* \param file_priv DRM file private.
* \return zero on success or a negative number on failure.
*
* If the hardware lock is held then free it, and take it again for the kernel
* context since it's necessary to reclaim buffers. Unlink the file private
* data from its list and free it. Decreases the open count and if it reaches
* zero calls drm_lastclose().
*/
int drm_release(struct inode *inode, struct file *filp)
{
struct drm_file *file_priv = filp->private_data;
struct drm_device *dev = file_priv->minor->dev;
int retcode = 0;
drm: Fix race that can lockup the kernel The i915_vblank_swap() function schedules an automatic buffer swap upon receipt of the vertical sync interrupt. Such an operation is lengthy so it can't be allowed to happen in normal interrupt context, thus the DRM implements this by scheduling the work in a kernel softirq-scheduled tasklet. In order for the buffer swap to work safely, the DRM's central lock must be taken, via a call to drm_lock_take() located in drivers/char/drm/drm_irq.c within the function drm_locked_tasklet_func(). The lock-taking logic uses a non-interrupt-blocking spinlock to implement the manipulations needed to take the lock. This semantic would be safe if all attempts to use the spinlock only happen from process context. However this buffer swap happens from softirq context which is really a form of interrupt context. Thus we have an unsafe situation, in that drm_locked_tasklet_func() can block on a spinlock already taken by a thread in process context which will never get scheduled again because of the blocked softirq tasklet. This wedges the kernel hard. To trigger this bug, run a dual-head cloned mode configuration which uses the i915 drm, then execute an opengl application which synchronizes buffer swaps against the vertical sync interrupt. In my testing, a lockup always results after running anywhere from 5 minutes to an hour and a half. I believe dual-head is needed to really trigger the problem because then the vertical sync interrupt handling is no longer predictable (due to being interrupt-sourced from two different heads running at different speeds). This raises the probability of the tasklet trying to run while the userspace DRI is doing things to the GPU (and manipulating the DRM lock). The fix is to change the relevant spinlock semantics to be the interrupt-blocking form. After this change I am no longer able to trigger the lockup; the longest test run so far was 20 hours (test stopped after that point). Note: I have examined the places where this spinlock is being employed; all are reasonably short bounded sequences and should be suitable for interrupts being blocked without impacting overall kernel interrupt response latency. Signed-off-by: Mike Isely <isely@pobox.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2008-03-14 03:30:35 +07:00
unsigned long irqflags;
lock_kernel();
DRM_DEBUG("open_count = %d\n", dev->open_count);
if (dev->driver->preclose)
dev->driver->preclose(dev, file_priv);
/* ========================================================
* Begin inline drm_release
*/
DRM_DEBUG("pid = %d, device = 0x%lx, open_count = %d\n",
task_pid_nr(current),
(long)old_encode_dev(file_priv->minor->device),
dev->open_count);
if (dev->driver->reclaim_buffers_locked && dev->lock.hw_lock) {
if (drm_i_have_hw_lock(dev, file_priv)) {
dev->driver->reclaim_buffers_locked(dev, file_priv);
} else {
unsigned long _end=jiffies + 3*DRM_HZ;
int locked = 0;
drm_idlelock_take(&dev->lock);
/*
* Wait for a while.
*/
do{
drm: Fix race that can lockup the kernel The i915_vblank_swap() function schedules an automatic buffer swap upon receipt of the vertical sync interrupt. Such an operation is lengthy so it can't be allowed to happen in normal interrupt context, thus the DRM implements this by scheduling the work in a kernel softirq-scheduled tasklet. In order for the buffer swap to work safely, the DRM's central lock must be taken, via a call to drm_lock_take() located in drivers/char/drm/drm_irq.c within the function drm_locked_tasklet_func(). The lock-taking logic uses a non-interrupt-blocking spinlock to implement the manipulations needed to take the lock. This semantic would be safe if all attempts to use the spinlock only happen from process context. However this buffer swap happens from softirq context which is really a form of interrupt context. Thus we have an unsafe situation, in that drm_locked_tasklet_func() can block on a spinlock already taken by a thread in process context which will never get scheduled again because of the blocked softirq tasklet. This wedges the kernel hard. To trigger this bug, run a dual-head cloned mode configuration which uses the i915 drm, then execute an opengl application which synchronizes buffer swaps against the vertical sync interrupt. In my testing, a lockup always results after running anywhere from 5 minutes to an hour and a half. I believe dual-head is needed to really trigger the problem because then the vertical sync interrupt handling is no longer predictable (due to being interrupt-sourced from two different heads running at different speeds). This raises the probability of the tasklet trying to run while the userspace DRI is doing things to the GPU (and manipulating the DRM lock). The fix is to change the relevant spinlock semantics to be the interrupt-blocking form. After this change I am no longer able to trigger the lockup; the longest test run so far was 20 hours (test stopped after that point). Note: I have examined the places where this spinlock is being employed; all are reasonably short bounded sequences and should be suitable for interrupts being blocked without impacting overall kernel interrupt response latency. Signed-off-by: Mike Isely <isely@pobox.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2008-03-14 03:30:35 +07:00
spin_lock_irqsave(&dev->lock.spinlock,
irqflags);
locked = dev->lock.idle_has_lock;
drm: Fix race that can lockup the kernel The i915_vblank_swap() function schedules an automatic buffer swap upon receipt of the vertical sync interrupt. Such an operation is lengthy so it can't be allowed to happen in normal interrupt context, thus the DRM implements this by scheduling the work in a kernel softirq-scheduled tasklet. In order for the buffer swap to work safely, the DRM's central lock must be taken, via a call to drm_lock_take() located in drivers/char/drm/drm_irq.c within the function drm_locked_tasklet_func(). The lock-taking logic uses a non-interrupt-blocking spinlock to implement the manipulations needed to take the lock. This semantic would be safe if all attempts to use the spinlock only happen from process context. However this buffer swap happens from softirq context which is really a form of interrupt context. Thus we have an unsafe situation, in that drm_locked_tasklet_func() can block on a spinlock already taken by a thread in process context which will never get scheduled again because of the blocked softirq tasklet. This wedges the kernel hard. To trigger this bug, run a dual-head cloned mode configuration which uses the i915 drm, then execute an opengl application which synchronizes buffer swaps against the vertical sync interrupt. In my testing, a lockup always results after running anywhere from 5 minutes to an hour and a half. I believe dual-head is needed to really trigger the problem because then the vertical sync interrupt handling is no longer predictable (due to being interrupt-sourced from two different heads running at different speeds). This raises the probability of the tasklet trying to run while the userspace DRI is doing things to the GPU (and manipulating the DRM lock). The fix is to change the relevant spinlock semantics to be the interrupt-blocking form. After this change I am no longer able to trigger the lockup; the longest test run so far was 20 hours (test stopped after that point). Note: I have examined the places where this spinlock is being employed; all are reasonably short bounded sequences and should be suitable for interrupts being blocked without impacting overall kernel interrupt response latency. Signed-off-by: Mike Isely <isely@pobox.com> Signed-off-by: Dave Airlie <airlied@redhat.com>
2008-03-14 03:30:35 +07:00
spin_unlock_irqrestore(&dev->lock.spinlock,
irqflags);
if (locked)
break;
schedule();
} while (!time_after_eq(jiffies, _end));
if (!locked) {
DRM_ERROR("reclaim_buffers_locked() deadlock. Please rework this\n"
"\tdriver to use reclaim_buffers_idlelocked() instead.\n"
"\tI will go on reclaiming the buffers anyway.\n");
}
dev->driver->reclaim_buffers_locked(dev, file_priv);
drm_idlelock_release(&dev->lock);
}
}
if (dev->driver->reclaim_buffers_idlelocked && dev->lock.hw_lock) {
drm_idlelock_take(&dev->lock);
dev->driver->reclaim_buffers_idlelocked(dev, file_priv);
drm_idlelock_release(&dev->lock);
}
if (drm_i_have_hw_lock(dev, file_priv)) {
DRM_DEBUG("File %p released, freeing lock for context %d\n",
filp, _DRM_LOCKING_CONTEXT(dev->lock.hw_lock->lock));
drm_lock_free(&dev->lock,
_DRM_LOCKING_CONTEXT(dev->lock.hw_lock->lock));
}
if (drm_core_check_feature(dev, DRIVER_HAVE_DMA) &&
!dev->driver->reclaim_buffers_locked) {
dev->driver->reclaim_buffers(dev, file_priv);
}
drm_fasync(-1, filp, 0);
mutex_lock(&dev->ctxlist_mutex);
if (!list_empty(&dev->ctxlist)) {
struct drm_ctx_list *pos, *n;
list_for_each_entry_safe(pos, n, &dev->ctxlist, head) {
if (pos->tag == file_priv &&
pos->handle != DRM_KERNEL_CONTEXT) {
if (dev->driver->context_dtor)
dev->driver->context_dtor(dev,
pos->handle);
drm_ctxbitmap_free(dev, pos->handle);
list_del(&pos->head);
drm_free(pos, sizeof(*pos), DRM_MEM_CTXLIST);
--dev->ctx_count;
}
}
}
mutex_unlock(&dev->ctxlist_mutex);
mutex_lock(&dev->struct_mutex);
if (file_priv->remove_auth_on_close == 1) {
struct drm_file *temp;
list_for_each_entry(temp, &dev->filelist, lhead)
temp->authenticated = 0;
}
list_del(&file_priv->lhead);
mutex_unlock(&dev->struct_mutex);
if (dev->driver->postclose)
dev->driver->postclose(dev, file_priv);
drm_free(file_priv, sizeof(*file_priv), DRM_MEM_FILES);
/* ========================================================
* End inline drm_release
*/
atomic_inc(&dev->counts[_DRM_STAT_CLOSES]);
spin_lock(&dev->count_lock);
if (!--dev->open_count) {
if (atomic_read(&dev->ioctl_count) || dev->blocked) {
DRM_ERROR("Device busy: %d %d\n",
atomic_read(&dev->ioctl_count), dev->blocked);
spin_unlock(&dev->count_lock);
unlock_kernel();
return -EBUSY;
}
spin_unlock(&dev->count_lock);
unlock_kernel();
return drm_lastclose(dev);
}
spin_unlock(&dev->count_lock);
unlock_kernel();
return retcode;
}
EXPORT_SYMBOL(drm_release);
/** No-op. */
unsigned int drm_poll(struct file *filp, struct poll_table_struct *wait)
{
return 0;
}
EXPORT_SYMBOL(drm_poll);