2005-04-16 22:20:36 +00:00
|
|
|
/* mga_irq.c -- IRQ handling for radeon -*- linux-c -*-
|
|
|
|
*
|
|
|
|
* Copyright (C) The Weather Channel, Inc. 2002. All Rights Reserved.
|
2005-09-25 04:28:13 +00:00
|
|
|
*
|
2005-04-16 22:20:36 +00:00
|
|
|
* The Weather Channel (TM) funded Tungsten Graphics to develop the
|
|
|
|
* initial release of the Radeon 8500 driver under the XFree86 license.
|
|
|
|
* This notice must be preserved.
|
|
|
|
*
|
|
|
|
* 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
|
|
|
|
* PRECISION INSIGHT 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.
|
|
|
|
*
|
|
|
|
* Authors:
|
|
|
|
* Keith Whitwell <keith@tungstengraphics.com>
|
|
|
|
* Eric Anholt <anholt@FreeBSD.org>
|
|
|
|
*/
|
|
|
|
|
|
|
|
#include "drmP.h"
|
|
|
|
#include "drm.h"
|
|
|
|
#include "mga_drm.h"
|
|
|
|
#include "mga_drv.h"
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
irqreturn_t mga_driver_irq_handler(DRM_IRQ_ARGS)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
|
|
|
drm_device_t *dev = (drm_device_t *) arg;
|
2005-09-25 04:28:13 +00:00
|
|
|
drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
|
2005-04-16 22:20:36 +00:00
|
|
|
int status;
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
int handled = 0;
|
|
|
|
|
|
|
|
status = MGA_READ(MGA_STATUS);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
/* VBLANK interrupt */
|
2005-09-25 04:28:13 +00:00
|
|
|
if (status & MGA_VLINEPEN) {
|
|
|
|
MGA_WRITE(MGA_ICLEAR, MGA_VLINEICLR);
|
2005-04-16 22:20:36 +00:00
|
|
|
atomic_inc(&dev->vbl_received);
|
|
|
|
DRM_WAKEUP(&dev->vbl_queue);
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
drm_vbl_send_signals(dev);
|
|
|
|
handled = 1;
|
|
|
|
}
|
|
|
|
|
|
|
|
/* SOFTRAP interrupt */
|
|
|
|
if (status & MGA_SOFTRAPEN) {
|
|
|
|
const u32 prim_start = MGA_READ(MGA_PRIMADDRESS);
|
2005-09-25 04:28:13 +00:00
|
|
|
const u32 prim_end = MGA_READ(MGA_PRIMEND);
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
|
|
|
|
MGA_WRITE(MGA_ICLEAR, MGA_SOFTRAPICLR);
|
|
|
|
|
|
|
|
/* In addition to clearing the interrupt-pending bit, we
|
|
|
|
* have to write to MGA_PRIMEND to re-start the DMA operation.
|
|
|
|
*/
|
2005-09-25 04:28:13 +00:00
|
|
|
if ((prim_start & ~0x03) != (prim_end & ~0x03)) {
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
MGA_WRITE(MGA_PRIMEND, prim_end);
|
|
|
|
}
|
|
|
|
|
|
|
|
atomic_inc(&dev_priv->last_fence_retired);
|
|
|
|
DRM_WAKEUP(&dev_priv->fence_queue);
|
|
|
|
handled = 1;
|
|
|
|
}
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
if (handled) {
|
2005-04-16 22:20:36 +00:00
|
|
|
return IRQ_HANDLED;
|
|
|
|
}
|
|
|
|
return IRQ_NONE;
|
|
|
|
}
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
int mga_driver_vblank_wait(drm_device_t * dev, unsigned int *sequence)
|
2005-04-16 22:20:36 +00:00
|
|
|
{
|
|
|
|
unsigned int cur_vblank;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/* Assume that the user has missed the current sequence number
|
|
|
|
* by about a day rather than she wants to wait for years
|
2005-09-25 04:28:13 +00:00
|
|
|
* using vertical blanks...
|
2005-04-16 22:20:36 +00:00
|
|
|
*/
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_WAIT_ON(ret, dev->vbl_queue, 3 * DRM_HZ,
|
|
|
|
(((cur_vblank = atomic_read(&dev->vbl_received))
|
|
|
|
- *sequence) <= (1 << 23)));
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
*sequence = cur_vblank;
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
int mga_driver_fence_wait(drm_device_t * dev, unsigned int *sequence)
|
|
|
|
{
|
|
|
|
drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
|
|
|
|
unsigned int cur_fence;
|
|
|
|
int ret = 0;
|
|
|
|
|
|
|
|
/* Assume that the user has missed the current sequence number
|
|
|
|
* by about a day rather than she wants to wait for years
|
|
|
|
* using fences.
|
|
|
|
*/
|
|
|
|
DRM_WAIT_ON(ret, dev_priv->fence_queue, 3 * DRM_HZ,
|
|
|
|
(((cur_fence = atomic_read(&dev_priv->last_fence_retired))
|
|
|
|
- *sequence) <= (1 << 23)));
|
|
|
|
|
|
|
|
*sequence = cur_fence;
|
|
|
|
|
|
|
|
return ret;
|
|
|
|
}
|
|
|
|
|
|
|
|
void mga_driver_irq_preinstall(drm_device_t * dev)
|
|
|
|
{
|
|
|
|
drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
|
2005-04-16 22:20:36 +00:00
|
|
|
|
|
|
|
/* Disable *all* interrupts */
|
2005-09-25 04:28:13 +00:00
|
|
|
MGA_WRITE(MGA_IEN, 0);
|
2005-04-16 22:20:36 +00:00
|
|
|
/* Clear bits if they're already high */
|
2005-09-25 04:28:13 +00:00
|
|
|
MGA_WRITE(MGA_ICLEAR, ~0);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
void mga_driver_irq_postinstall(drm_device_t * dev)
|
|
|
|
{
|
|
|
|
drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
DRM_INIT_WAITQUEUE(&dev_priv->fence_queue);
|
2005-04-16 22:20:36 +00:00
|
|
|
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
/* Turn on vertical blank interrupt and soft trap interrupt. */
|
|
|
|
MGA_WRITE(MGA_IEN, MGA_VLINEIEN | MGA_SOFTRAPEN);
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|
|
|
|
|
2005-09-25 04:28:13 +00:00
|
|
|
void mga_driver_irq_uninstall(drm_device_t * dev)
|
|
|
|
{
|
|
|
|
drm_mga_private_t *dev_priv = (drm_mga_private_t *) dev->dev_private;
|
2005-04-16 22:20:36 +00:00
|
|
|
if (!dev_priv)
|
|
|
|
return;
|
|
|
|
|
|
|
|
/* Disable *all* interrupts */
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
MGA_WRITE(MGA_IEN, 0);
|
2005-09-25 04:28:13 +00:00
|
|
|
|
Add support for PCI MGA cards to MGA DRM.
This patch adds serveral new ioctls and a new query to get_param query to
support PCI MGA cards.
Two ioctls were added to implement interrupt based waiting. With this change,
the client-side driver no longer needs to map the primary DMA region or the
MMIO region. Previously, end-of-frame waiting was done by busy waiting in the
client-side driver until one of the MMIO registers (the current DMA pointer)
matched a pointer to the end of primary DMA space. By using interrupts, the
busy waiting and the extra mappings are removed.
A third ioctl was added to bootstrap DMA. This ioctl, which is used by the
X-server, moves a *LOT* of code from the X-server into the kernel. This allows
the kernel to do whatever needs to be done to setup DMA buffers. The entire
process and the locations of the buffers are hidden from user-mode.
Additionally, a get_param query was added to differentiate between G4x0 cards
and G550 cards. A gap was left in the numbering sequence so that, if needed,
G450 cards could be distinguished from G400 cards. According to Ville
Syrjälä, the G4x0 cards and the G550 cards handle anisotropic filtering
differently. This seems the most compatible way to let the client-side driver
know which card it's own. Doing this very small change now eliminates the
need to bump the DRM minor version twice.
http://marc.theaimsgroup.com/?l=dri-devel&m=106625815319773&w=2
(airlied - this may not work at this point, I think the follow on buffer
cleanup patches will be needed)
From: Ian Romanick <idr@us.ibm.com>
Signed-off-by: Dave Airlie <airlied@linux.ie>
2005-07-10 08:20:09 +00:00
|
|
|
dev->irq_enabled = 0;
|
2005-04-16 22:20:36 +00:00
|
|
|
}
|