kernel_optimize_test/drivers/video/nvidia/nv_accel.c
Dave Jones e15de77e74 nvidiafb: prevent triggering of softlockup
If the chip locks up, we get into a long polling loop,
where the softlockup detector kicks in.
See https://bugzilla.redhat.com/bugzilla/attachment.cgi?id=151878
for an example.

[adaplas]
Chip lockup can occur at 3 points (flush, sync, and wait). Consolidate and
allow the driver to go to safe mode cleanly.

Signed-off-by: Dave Jones <davej@redhat.com>
Signed-off-by: Antonino Daplas <adaplas@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2007-05-08 11:15:32 -07:00

413 lines
12 KiB
C

/***************************************************************************\
|* *|
|* Copyright 1993-2003 NVIDIA, Corporation. All rights reserved. *|
|* *|
|* NOTICE TO USER: The source code is copyrighted under U.S. and *|
|* international laws. Users and possessors of this source code are *|
|* hereby granted a nonexclusive, royalty-free copyright license to *|
|* use this code in individual and commercial software. *|
|* *|
|* Any use of this source code must include, in the user documenta- *|
|* tion and internal comments to the code, notices to the end user *|
|* as follows: *|
|* *|
|* Copyright 1993-2003 NVIDIA, Corporation. All rights reserved. *|
|* *|
|* NVIDIA, CORPORATION MAKES NO REPRESENTATION ABOUT THE SUITABILITY *|
|* OF THIS SOURCE CODE FOR ANY PURPOSE. IT IS PROVIDED "AS IS" *|
|* WITHOUT EXPRESS OR IMPLIED WARRANTY OF ANY KIND. NVIDIA, CORPOR- *|
|* ATION DISCLAIMS ALL WARRANTIES WITH REGARD TO THIS SOURCE CODE, *|
|* INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY, NONINFRINGE- *|
|* MENT, AND FITNESS FOR A PARTICULAR PURPOSE. IN NO EVENT SHALL *|
|* NVIDIA, CORPORATION BE LIABLE FOR ANY SPECIAL, INDIRECT, INCI- *|
|* DENTAL, OR CONSEQUENTIAL DAMAGES, OR ANY DAMAGES WHATSOEVER RE- *|
|* SULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN ACTION *|
|* OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF *|
|* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOURCE CODE. *|
|* *|
|* U.S. Government End Users. This source code is a "commercial *|
|* item," as that term is defined at 48 C.F.R. 2.101 (OCT 1995), *|
|* consisting of "commercial computer software" and "commercial *|
|* computer software documentation," as such terms are used in *|
|* 48 C.F.R. 12.212 (SEPT 1995) and is provided to the U.S. Govern- *|
|* ment only as a commercial end item. Consistent with 48 C.F.R. *|
|* 12.212 and 48 C.F.R. 227.7202-1 through 227.7202-4 (JUNE 1995), *|
|* all U.S. Government End Users acquire the source code with only *|
|* those rights set forth herein. *|
|* *|
\***************************************************************************/
/*
* GPL Licensing Note - According to Mark Vojkovich, author of the Xorg/
* XFree86 'nv' driver, this source code is provided under MIT-style licensing
* where the source code is provided "as is" without warranty of any kind.
* The only usage restriction is for the copyright notices to be retained
* whenever code is used.
*
* Antonino Daplas <adaplas@pol.net> 2005-03-11
*/
#include <linux/fb.h>
#include "nv_type.h"
#include "nv_proto.h"
#include "nv_dma.h"
#include "nv_local.h"
/* There is a HW race condition with videoram command buffers.
You can't jump to the location of your put offset. We write put
at the jump offset + SKIPS dwords with noop padding in between
to solve this problem */
#define SKIPS 8
static const int NVCopyROP[16] = {
0xCC, /* copy */
0x55 /* invert */
};
static const int NVCopyROP_PM[16] = {
0xCA, /* copy */
0x5A, /* invert */
};
static inline void nvidiafb_safe_mode(struct fb_info *info)
{
struct nvidia_par *par = info->par;
touch_softlockup_watchdog();
info->pixmap.scan_align = 1;
par->lockup = 1;
}
static inline void NVFlush(struct fb_info *info)
{
struct nvidia_par *par = info->par;
int count = 1000000000;
while (--count && READ_GET(par) != par->dmaPut) ;
if (!count) {
printk("nvidiafb: DMA Flush lockup\n");
nvidiafb_safe_mode(info);
}
}
static inline void NVSync(struct fb_info *info)
{
struct nvidia_par *par = info->par;
int count = 1000000000;
while (--count && NV_RD32(par->PGRAPH, 0x0700)) ;
if (!count) {
printk("nvidiafb: DMA Sync lockup\n");
nvidiafb_safe_mode(info);
}
}
static void NVDmaKickoff(struct nvidia_par *par)
{
if (par->dmaCurrent != par->dmaPut) {
par->dmaPut = par->dmaCurrent;
WRITE_PUT(par, par->dmaPut);
}
}
static void NVDmaWait(struct fb_info *info, int size)
{
struct nvidia_par *par = info->par;
int dmaGet;
int count = 1000000000, cnt;
size++;
while (par->dmaFree < size && --count && !par->lockup) {
dmaGet = READ_GET(par);
if (par->dmaPut >= dmaGet) {
par->dmaFree = par->dmaMax - par->dmaCurrent;
if (par->dmaFree < size) {
NVDmaNext(par, 0x20000000);
if (dmaGet <= SKIPS) {
if (par->dmaPut <= SKIPS)
WRITE_PUT(par, SKIPS + 1);
cnt = 1000000000;
do {
dmaGet = READ_GET(par);
} while (--cnt && dmaGet <= SKIPS);
if (!cnt) {
printk("DMA Get lockup\n");
par->lockup = 1;
}
}
WRITE_PUT(par, SKIPS);
par->dmaCurrent = par->dmaPut = SKIPS;
par->dmaFree = dmaGet - (SKIPS + 1);
}
} else
par->dmaFree = dmaGet - par->dmaCurrent - 1;
}
if (!count) {
printk("nvidiafb: DMA Wait Lockup\n");
nvidiafb_safe_mode(info);
}
}
static void NVSetPattern(struct fb_info *info, u32 clr0, u32 clr1,
u32 pat0, u32 pat1)
{
struct nvidia_par *par = info->par;
NVDmaStart(info, par, PATTERN_COLOR_0, 4);
NVDmaNext(par, clr0);
NVDmaNext(par, clr1);
NVDmaNext(par, pat0);
NVDmaNext(par, pat1);
}
static void NVSetRopSolid(struct fb_info *info, u32 rop, u32 planemask)
{
struct nvidia_par *par = info->par;
if (planemask != ~0) {
NVSetPattern(info, 0, planemask, ~0, ~0);
if (par->currentRop != (rop + 32)) {
NVDmaStart(info, par, ROP_SET, 1);
NVDmaNext(par, NVCopyROP_PM[rop]);
par->currentRop = rop + 32;
}
} else if (par->currentRop != rop) {
if (par->currentRop >= 16)
NVSetPattern(info, ~0, ~0, ~0, ~0);
NVDmaStart(info, par, ROP_SET, 1);
NVDmaNext(par, NVCopyROP[rop]);
par->currentRop = rop;
}
}
static void NVSetClippingRectangle(struct fb_info *info, int x1, int y1,
int x2, int y2)
{
struct nvidia_par *par = info->par;
int h = y2 - y1 + 1;
int w = x2 - x1 + 1;
NVDmaStart(info, par, CLIP_POINT, 2);
NVDmaNext(par, (y1 << 16) | x1);
NVDmaNext(par, (h << 16) | w);
}
void NVResetGraphics(struct fb_info *info)
{
struct nvidia_par *par = info->par;
u32 surfaceFormat, patternFormat, rectFormat, lineFormat;
int pitch, i;
pitch = info->fix.line_length;
par->dmaBase = (u32 __iomem *) (&par->FbStart[par->FbUsableSize]);
for (i = 0; i < SKIPS; i++)
NV_WR32(&par->dmaBase[i], 0, 0x00000000);
NV_WR32(&par->dmaBase[0x0 + SKIPS], 0, 0x00040000);
NV_WR32(&par->dmaBase[0x1 + SKIPS], 0, 0x80000010);
NV_WR32(&par->dmaBase[0x2 + SKIPS], 0, 0x00042000);
NV_WR32(&par->dmaBase[0x3 + SKIPS], 0, 0x80000011);
NV_WR32(&par->dmaBase[0x4 + SKIPS], 0, 0x00044000);
NV_WR32(&par->dmaBase[0x5 + SKIPS], 0, 0x80000012);
NV_WR32(&par->dmaBase[0x6 + SKIPS], 0, 0x00046000);
NV_WR32(&par->dmaBase[0x7 + SKIPS], 0, 0x80000013);
NV_WR32(&par->dmaBase[0x8 + SKIPS], 0, 0x00048000);
NV_WR32(&par->dmaBase[0x9 + SKIPS], 0, 0x80000014);
NV_WR32(&par->dmaBase[0xA + SKIPS], 0, 0x0004A000);
NV_WR32(&par->dmaBase[0xB + SKIPS], 0, 0x80000015);
NV_WR32(&par->dmaBase[0xC + SKIPS], 0, 0x0004C000);
NV_WR32(&par->dmaBase[0xD + SKIPS], 0, 0x80000016);
NV_WR32(&par->dmaBase[0xE + SKIPS], 0, 0x0004E000);
NV_WR32(&par->dmaBase[0xF + SKIPS], 0, 0x80000017);
par->dmaPut = 0;
par->dmaCurrent = 16 + SKIPS;
par->dmaMax = 8191;
par->dmaFree = par->dmaMax - par->dmaCurrent;
switch (info->var.bits_per_pixel) {
case 32:
case 24:
surfaceFormat = SURFACE_FORMAT_DEPTH24;
patternFormat = PATTERN_FORMAT_DEPTH24;
rectFormat = RECT_FORMAT_DEPTH24;
lineFormat = LINE_FORMAT_DEPTH24;
break;
case 16:
surfaceFormat = SURFACE_FORMAT_DEPTH16;
patternFormat = PATTERN_FORMAT_DEPTH16;
rectFormat = RECT_FORMAT_DEPTH16;
lineFormat = LINE_FORMAT_DEPTH16;
break;
default:
surfaceFormat = SURFACE_FORMAT_DEPTH8;
patternFormat = PATTERN_FORMAT_DEPTH8;
rectFormat = RECT_FORMAT_DEPTH8;
lineFormat = LINE_FORMAT_DEPTH8;
break;
}
NVDmaStart(info, par, SURFACE_FORMAT, 4);
NVDmaNext(par, surfaceFormat);
NVDmaNext(par, pitch | (pitch << 16));
NVDmaNext(par, 0);
NVDmaNext(par, 0);
NVDmaStart(info, par, PATTERN_FORMAT, 1);
NVDmaNext(par, patternFormat);
NVDmaStart(info, par, RECT_FORMAT, 1);
NVDmaNext(par, rectFormat);
NVDmaStart(info, par, LINE_FORMAT, 1);
NVDmaNext(par, lineFormat);
par->currentRop = ~0; /* set to something invalid */
NVSetRopSolid(info, ROP_COPY, ~0);
NVSetClippingRectangle(info, 0, 0, info->var.xres_virtual,
info->var.yres_virtual);
NVDmaKickoff(par);
}
int nvidiafb_sync(struct fb_info *info)
{
struct nvidia_par *par = info->par;
if (info->state != FBINFO_STATE_RUNNING)
return 0;
if (!par->lockup)
NVFlush(info);
if (!par->lockup)
NVSync(info);
return 0;
}
void nvidiafb_copyarea(struct fb_info *info, const struct fb_copyarea *region)
{
struct nvidia_par *par = info->par;
if (info->state != FBINFO_STATE_RUNNING)
return;
if (par->lockup)
return cfb_copyarea(info, region);
NVDmaStart(info, par, BLIT_POINT_SRC, 3);
NVDmaNext(par, (region->sy << 16) | region->sx);
NVDmaNext(par, (region->dy << 16) | region->dx);
NVDmaNext(par, (region->height << 16) | region->width);
NVDmaKickoff(par);
}
void nvidiafb_fillrect(struct fb_info *info, const struct fb_fillrect *rect)
{
struct nvidia_par *par = info->par;
u32 color;
if (info->state != FBINFO_STATE_RUNNING)
return;
if (par->lockup)
return cfb_fillrect(info, rect);
if (info->var.bits_per_pixel == 8)
color = rect->color;
else
color = ((u32 *) info->pseudo_palette)[rect->color];
if (rect->rop != ROP_COPY)
NVSetRopSolid(info, rect->rop, ~0);
NVDmaStart(info, par, RECT_SOLID_COLOR, 1);
NVDmaNext(par, color);
NVDmaStart(info, par, RECT_SOLID_RECTS(0), 2);
NVDmaNext(par, (rect->dx << 16) | rect->dy);
NVDmaNext(par, (rect->width << 16) | rect->height);
NVDmaKickoff(par);
if (rect->rop != ROP_COPY)
NVSetRopSolid(info, ROP_COPY, ~0);
}
static void nvidiafb_mono_color_expand(struct fb_info *info,
const struct fb_image *image)
{
struct nvidia_par *par = info->par;
u32 fg, bg, mask = ~(~0 >> (32 - info->var.bits_per_pixel));
u32 dsize, width, *data = (u32 *) image->data, tmp;
int j, k = 0;
width = (image->width + 31) & ~31;
dsize = (width * image->height) >> 5;
if (info->var.bits_per_pixel == 8) {
fg = image->fg_color | mask;
bg = image->bg_color | mask;
} else {
fg = ((u32 *) info->pseudo_palette)[image->fg_color] | mask;
bg = ((u32 *) info->pseudo_palette)[image->bg_color] | mask;
}
NVDmaStart(info, par, RECT_EXPAND_TWO_COLOR_CLIP, 7);
NVDmaNext(par, (image->dy << 16) | (image->dx & 0xffff));
NVDmaNext(par, ((image->dy + image->height) << 16) |
((image->dx + image->width) & 0xffff));
NVDmaNext(par, bg);
NVDmaNext(par, fg);
NVDmaNext(par, (image->height << 16) | width);
NVDmaNext(par, (image->height << 16) | width);
NVDmaNext(par, (image->dy << 16) | (image->dx & 0xffff));
while (dsize >= RECT_EXPAND_TWO_COLOR_DATA_MAX_DWORDS) {
NVDmaStart(info, par, RECT_EXPAND_TWO_COLOR_DATA(0),
RECT_EXPAND_TWO_COLOR_DATA_MAX_DWORDS);
for (j = RECT_EXPAND_TWO_COLOR_DATA_MAX_DWORDS; j--;) {
tmp = data[k++];
reverse_order(&tmp);
NVDmaNext(par, tmp);
}
dsize -= RECT_EXPAND_TWO_COLOR_DATA_MAX_DWORDS;
}
if (dsize) {
NVDmaStart(info, par, RECT_EXPAND_TWO_COLOR_DATA(0), dsize);
for (j = dsize; j--;) {
tmp = data[k++];
reverse_order(&tmp);
NVDmaNext(par, tmp);
}
}
NVDmaKickoff(par);
}
void nvidiafb_imageblit(struct fb_info *info, const struct fb_image *image)
{
struct nvidia_par *par = info->par;
if (info->state != FBINFO_STATE_RUNNING)
return;
if (image->depth == 1 && !par->lockup)
nvidiafb_mono_color_expand(info, image);
else
cfb_imageblit(info, image);
}