Minki/linux
1
0
Fork 1
mirror of https://github.com/torvalds/linux.git synced 2024-11-08 21:21:47 +00:00
Code Issues Packages Projects Releases Wiki Activity
332b242f47
linux/drivers/gpu/drm/nouveau/nv40_fifo.c
Francisco Jerez 3945e47543 drm/nouveau: Refactor context destruction to avoid a lock ordering issue. 
The destroy_context() engine hooks call gpuobj management functions to
release the channel resources, these functions use HARDIRQ-unsafe locks
whereas destroy_context() is called with the HARDIRQ-safe
context_switch_lock held, that's a lock ordering violation.

Push the engine-specific channel destruction logic into destroy_context()
and let the hardware-specific code lock and unlock when it's actually
needed. Change the engine destruction order to avoid a race in the small
gap between pgraph and pfifo context uninitialization.

Reported-by: Marcin Slusarz <marcin.slusarz@gmail.com>
Signed-off-by: Francisco Jerez <currojerez@riseup.net>
Signed-off-by: Ben Skeggs <bskeggs@redhat.com>
2010-12-03 15:06:35 +10:00

301 lines
9.7 KiB
C
Raw Blame History

/*
* Copyright (C) 2007 Ben Skeggs.
* 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 THE COPYRIGHT OWNER(S) 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 "nouveau_drv.h"
#include "nouveau_drm.h"
#include "nouveau_ramht.h"
#define NV40_RAMFC(c) (dev_priv->ramfc->pinst + ((c) * NV40_RAMFC__SIZE))
#define NV40_RAMFC__SIZE 128
int
nv40_fifo_create_context(struct nouveau_channel *chan)
{
struct drm_device *dev = chan->dev;
struct drm_nouveau_private *dev_priv = dev->dev_private;
uint32_t fc = NV40_RAMFC(chan->id);
unsigned long flags;
int ret;
ret = nouveau_gpuobj_new_fake(dev, NV40_RAMFC(chan->id), ~0,
NV40_RAMFC__SIZE, NVOBJ_FLAG_ZERO_ALLOC |
NVOBJ_FLAG_ZERO_FREE, &chan->ramfc);
if (ret)
return ret;
spin_lock_irqsave(&dev_priv->context_switch_lock, flags);
nv_wi32(dev, fc + 0, chan->pushbuf_base);
nv_wi32(dev, fc + 4, chan->pushbuf_base);
nv_wi32(dev, fc + 12, chan->pushbuf->pinst >> 4);
nv_wi32(dev, fc + 24, NV_PFIFO_CACHE1_DMA_FETCH_TRIG_128_BYTES |
NV_PFIFO_CACHE1_DMA_FETCH_SIZE_128_BYTES |
NV_PFIFO_CACHE1_DMA_FETCH_MAX_REQS_8 |
#ifdef __BIG_ENDIAN
NV_PFIFO_CACHE1_BIG_ENDIAN |
#endif
0x30000000 /* no idea.. */);
nv_wi32(dev, fc + 56, chan->ramin_grctx->pinst >> 4);
nv_wi32(dev, fc + 60, 0x0001FFFF);
/* enable the fifo dma operation */
nv_wr32(dev, NV04_PFIFO_MODE,
nv_rd32(dev, NV04_PFIFO_MODE) | (1 << chan->id));
spin_unlock_irqrestore(&dev_priv->context_switch_lock, flags);
return 0;
}
static void
nv40_fifo_do_load_context(struct drm_device *dev, int chid)
{
struct drm_nouveau_private *dev_priv = dev->dev_private;
uint32_t fc = NV40_RAMFC(chid), tmp, tmp2;
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_PUT, nv_ri32(dev, fc + 0));
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_GET, nv_ri32(dev, fc + 4));
nv_wr32(dev, NV10_PFIFO_CACHE1_REF_CNT, nv_ri32(dev, fc + 8));
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_INSTANCE, nv_ri32(dev, fc + 12));
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_DCOUNT, nv_ri32(dev, fc + 16));
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_STATE, nv_ri32(dev, fc + 20));
/* No idea what 0x2058 is.. */
tmp = nv_ri32(dev, fc + 24);
tmp2 = nv_rd32(dev, 0x2058) & 0xFFF;
tmp2 |= (tmp & 0x30000000);
nv_wr32(dev, 0x2058, tmp2);
tmp &= ~0x30000000;
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_FETCH, tmp);
nv_wr32(dev, NV04_PFIFO_CACHE1_ENGINE, nv_ri32(dev, fc + 28));
nv_wr32(dev, NV04_PFIFO_CACHE1_PULL1, nv_ri32(dev, fc + 32));
nv_wr32(dev, NV10_PFIFO_CACHE1_ACQUIRE_VALUE, nv_ri32(dev, fc + 36));
tmp = nv_ri32(dev, fc + 40);
nv_wr32(dev, NV10_PFIFO_CACHE1_ACQUIRE_TIMESTAMP, tmp);
nv_wr32(dev, NV10_PFIFO_CACHE1_ACQUIRE_TIMEOUT, nv_ri32(dev, fc + 44));
nv_wr32(dev, NV10_PFIFO_CACHE1_SEMAPHORE, nv_ri32(dev, fc + 48));
nv_wr32(dev, NV10_PFIFO_CACHE1_DMA_SUBROUTINE, nv_ri32(dev, fc + 52));
nv_wr32(dev, NV40_PFIFO_GRCTX_INSTANCE, nv_ri32(dev, fc + 56));
/* Don't clobber the TIMEOUT_ENABLED flag when restoring from RAMFC */
tmp = nv_rd32(dev, NV04_PFIFO_DMA_TIMESLICE) & ~0x1FFFF;
tmp |= nv_ri32(dev, fc + 60) & 0x1FFFF;
nv_wr32(dev, NV04_PFIFO_DMA_TIMESLICE, tmp);
nv_wr32(dev, 0x32e4, nv_ri32(dev, fc + 64));
/* NVIDIA does this next line twice... */
nv_wr32(dev, 0x32e8, nv_ri32(dev, fc + 68));
nv_wr32(dev, 0x2088, nv_ri32(dev, fc + 76));
nv_wr32(dev, 0x3300, nv_ri32(dev, fc + 80));
nv_wr32(dev, NV03_PFIFO_CACHE1_GET, 0);
nv_wr32(dev, NV03_PFIFO_CACHE1_PUT, 0);
}
int
nv40_fifo_load_context(struct nouveau_channel *chan)
{
struct drm_device *dev = chan->dev;
uint32_t tmp;
nv40_fifo_do_load_context(dev, chan->id);
/* Set channel active, and in DMA mode */
nv_wr32(dev, NV03_PFIFO_CACHE1_PUSH1,
NV40_PFIFO_CACHE1_PUSH1_DMA | chan->id);
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_PUSH, 1);
/* Reset DMA_CTL_AT_INFO to INVALID */
tmp = nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_CTL) & ~(1 << 31);
nv_wr32(dev, NV04_PFIFO_CACHE1_DMA_CTL, tmp);
return 0;
}
int
nv40_fifo_unload_context(struct drm_device *dev)
{
struct drm_nouveau_private *dev_priv = dev->dev_private;
struct nouveau_fifo_engine *pfifo = &dev_priv->engine.fifo;
uint32_t fc, tmp;
int chid;
chid = pfifo->channel_id(dev);
if (chid < 0 || chid >= dev_priv->engine.fifo.channels)
return 0;
fc = NV40_RAMFC(chid);
nv_wi32(dev, fc + 0, nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_PUT));
nv_wi32(dev, fc + 4, nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_GET));
nv_wi32(dev, fc + 8, nv_rd32(dev, NV10_PFIFO_CACHE1_REF_CNT));
nv_wi32(dev, fc + 12, nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_INSTANCE));
nv_wi32(dev, fc + 16, nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_DCOUNT));
nv_wi32(dev, fc + 20, nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_STATE));
tmp = nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_FETCH);
tmp |= nv_rd32(dev, 0x2058) & 0x30000000;
nv_wi32(dev, fc + 24, tmp);
nv_wi32(dev, fc + 28, nv_rd32(dev, NV04_PFIFO_CACHE1_ENGINE));
nv_wi32(dev, fc + 32, nv_rd32(dev, NV04_PFIFO_CACHE1_PULL1));
nv_wi32(dev, fc + 36, nv_rd32(dev, NV10_PFIFO_CACHE1_ACQUIRE_VALUE));
tmp = nv_rd32(dev, NV10_PFIFO_CACHE1_ACQUIRE_TIMESTAMP);
nv_wi32(dev, fc + 40, tmp);
nv_wi32(dev, fc + 44, nv_rd32(dev, NV10_PFIFO_CACHE1_ACQUIRE_TIMEOUT));
nv_wi32(dev, fc + 48, nv_rd32(dev, NV10_PFIFO_CACHE1_SEMAPHORE));
/* NVIDIA read 0x3228 first, then write DMA_GET here.. maybe something
* more involved depending on the value of 0x3228?
*/
nv_wi32(dev, fc + 52, nv_rd32(dev, NV04_PFIFO_CACHE1_DMA_GET));
nv_wi32(dev, fc + 56, nv_rd32(dev, NV40_PFIFO_GRCTX_INSTANCE));
nv_wi32(dev, fc + 60, nv_rd32(dev, NV04_PFIFO_DMA_TIMESLICE) & 0x1ffff);
/* No idea what the below is for exactly, ripped from a mmio-trace */
nv_wi32(dev, fc + 64, nv_rd32(dev, NV40_PFIFO_UNK32E4));
/* NVIDIA do this next line twice.. bug? */
nv_wi32(dev, fc + 68, nv_rd32(dev, 0x32e8));
nv_wi32(dev, fc + 76, nv_rd32(dev, 0x2088));
nv_wi32(dev, fc + 80, nv_rd32(dev, 0x3300));
#if 0 /* no real idea which is PUT/GET in UNK_48.. */
tmp = nv_rd32(dev, NV04_PFIFO_CACHE1_GET);
tmp |= (nv_rd32(dev, NV04_PFIFO_CACHE1_PUT) << 16);
nv_wi32(dev, fc + 72, tmp);
#endif
nv40_fifo_do_load_context(dev, pfifo->channels - 1);
nv_wr32(dev, NV03_PFIFO_CACHE1_PUSH1,
NV40_PFIFO_CACHE1_PUSH1_DMA | (pfifo->channels - 1));
return 0;
}
static void
nv40_fifo_init_reset(struct drm_device *dev)
{
int i;
nv_wr32(dev, NV03_PMC_ENABLE,
nv_rd32(dev, NV03_PMC_ENABLE) & ~NV_PMC_ENABLE_PFIFO);
nv_wr32(dev, NV03_PMC_ENABLE,
nv_rd32(dev, NV03_PMC_ENABLE) | NV_PMC_ENABLE_PFIFO);
nv_wr32(dev, 0x003224, 0x000f0078);
nv_wr32(dev, 0x003210, 0x00000000);
nv_wr32(dev, 0x003270, 0x00000000);
nv_wr32(dev, 0x003240, 0x00000000);
nv_wr32(dev, 0x003244, 0x00000000);
nv_wr32(dev, 0x003258, 0x00000000);
nv_wr32(dev, 0x002504, 0x00000000);
for (i = 0; i < 16; i++)
nv_wr32(dev, 0x002510 + (i * 4), 0x00000000);
nv_wr32(dev, 0x00250c, 0x0000ffff);
nv_wr32(dev, 0x002048, 0x00000000);
nv_wr32(dev, 0x003228, 0x00000000);
nv_wr32(dev, 0x0032e8, 0x00000000);
nv_wr32(dev, 0x002410, 0x00000000);
nv_wr32(dev, 0x002420, 0x00000000);
nv_wr32(dev, 0x002058, 0x00000001);
nv_wr32(dev, 0x00221c, 0x00000000);
/* something with 0x2084, read/modify/write, no change */
nv_wr32(dev, 0x002040, 0x000000ff);
nv_wr32(dev, 0x002500, 0x00000000);
nv_wr32(dev, 0x003200, 0x00000000);
nv_wr32(dev, NV04_PFIFO_DMA_TIMESLICE, 0x2101ffff);
}
static void
nv40_fifo_init_ramxx(struct drm_device *dev)
{
struct drm_nouveau_private *dev_priv = dev->dev_private;
nv_wr32(dev, NV03_PFIFO_RAMHT, (0x03 << 24) /* search 128 */ |
((dev_priv->ramht->bits - 9) << 16) |
(dev_priv->ramht->gpuobj->pinst >> 8));
nv_wr32(dev, NV03_PFIFO_RAMRO, dev_priv->ramro->pinst >> 8);
switch (dev_priv->chipset) {
case 0x47:
case 0x49:
case 0x4b:
nv_wr32(dev, 0x2230, 1);
break;
default:
break;
}
switch (dev_priv->chipset) {
case 0x40:
case 0x41:
case 0x42:
case 0x43:
case 0x45:
case 0x47:
case 0x48:
case 0x49:
case 0x4b:
nv_wr32(dev, NV40_PFIFO_RAMFC, 0x30002);
break;
default:
nv_wr32(dev, 0x2230, 0);
nv_wr32(dev, NV40_PFIFO_RAMFC,
((dev_priv->vram_size - 512 * 1024 +
dev_priv->ramfc->pinst) >> 16) | (3 << 16));
break;
}
}
static void
nv40_fifo_init_intr(struct drm_device *dev)
{
nv_wr32(dev, 0x002100, 0xffffffff);
nv_wr32(dev, 0x002140, 0xffffffff);
}
int
nv40_fifo_init(struct drm_device *dev)
{
struct drm_nouveau_private *dev_priv = dev->dev_private;
struct nouveau_fifo_engine *pfifo = &dev_priv->engine.fifo;
int i;
nv40_fifo_init_reset(dev);
nv40_fifo_init_ramxx(dev);
nv40_fifo_do_load_context(dev, pfifo->channels - 1);
nv_wr32(dev, NV03_PFIFO_CACHE1_PUSH1, pfifo->channels - 1);
nv40_fifo_init_intr(dev);
pfifo->enable(dev);
pfifo->reassign(dev, true);
for (i = 0; i < dev_priv->engine.fifo.channels; i++) {
if (dev_priv->channels.ptr[i]) {
uint32_t mode = nv_rd32(dev, NV04_PFIFO_MODE);
nv_wr32(dev, NV04_PFIFO_MODE, mode | (1 << i));
}
}
return 0;
}
Reference in New Issue View Git Blame Copy Permalink