linux/drivers/video/fbdev/vesafb.c

498 lines
14 KiB
C
Raw Normal View History

/*
* framebuffer driver for VBE 2.0 compliant graphic boards
*
* switching to graphics mode happens at boot time (while
* running in real mode, see arch/i386/boot/video.S).
*
* (c) 1998 Gerd Knorr <kraxel@goldbach.in-berlin.de>
*
*/
#include <linux/module.h>
#include <linux/kernel.h>
#include <linux/errno.h>
#include <linux/string.h>
#include <linux/mm.h>
#include <linux/delay.h>
#include <linux/fb.h>
#include <linux/ioport.h>
#include <linux/init.h>
#include <linux/platform_device.h>
#include <linux/screen_info.h>
video: fbdev: vesafb: use arch_phys_wc_add() This driver uses the same area for MTRR as for the ioremap_wc(), if anything it just uses a smaller size in case MTRR reservation fails. ioremap_wc() API is already used to take advantage of architecture write-combining when available. Convert the driver from using the x86 specific MTRR code to the architecture agnostic arch_phys_wc_add(). arch_phys_wc_add() will avoid MTRR if write-combining is available. There are a few motivations for this: a) Take advantage of PAT when available b) Help bury MTRR code away, MTRR is architecture specific and on x86 its replaced by PAT c) Help with the goal of eventually using _PAGE_CACHE_UC over _PAGE_CACHE_UC_MINUS on x86 on ioremap_nocache() (see commit de33c442e titled "x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()") The conversion done is expressed by the following Coccinelle SmPL patch, it additionally required manual intervention to address all the #ifdery and removal of redundant things which arch_phys_wc_add() already addresses such as verbose message about when MTRR fails and doing nothing when we didn't get an MTRR. @ mtrr_found @ expression index, base, size; @@ -index = mtrr_add(base, size, MTRR_TYPE_WRCOMB, 1); +index = arch_phys_wc_add(base, size); @ mtrr_rm depends on mtrr_found @ expression mtrr_found.index, mtrr_found.base, mtrr_found.size; @@ -mtrr_del(index, base, size); +arch_phys_wc_del(index); @ mtrr_rm_zero_arg depends on mtrr_found @ expression mtrr_found.index; @@ -mtrr_del(index, 0, 0); +arch_phys_wc_del(index); @ mtrr_rm_fb_info depends on mtrr_found @ struct fb_info *info; expression mtrr_found.index; @@ -mtrr_del(index, info->fix.smem_start, info->fix.smem_len); +arch_phys_wc_del(index); @ ioremap_replace_nocache depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap_nocache(base, size); +info->screen_base = ioremap_wc(base, size); @ ioremap_replace_default depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap(base, size); +info->screen_base = ioremap_wc(base, size); Generated-by: Coccinelle SmPL Cc: Toshi Kani <toshi.kani@hp.com> Cc: Suresh Siddha <sbsiddha@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Juergen Gross <jgross@suse.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Dave Airlie <airlied@redhat.com> Cc: Antonino Daplas <adaplas@gmail.com> Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Jingoo Han <jg1.han@samsung.com> Cc: Wolfram Sang <wsa@the-dreams.de> Cc: Tomi Valkeinen <tomi.valkeinen@ti.com> Cc: linux-fbdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Luis R. Rodriguez <mcgrof@suse.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
2015-06-04 16:44:48 +00:00
#include <linux/io.h>
#include <video/vga.h>
#define dac_reg (0x3c8)
#define dac_val (0x3c9)
/* --------------------------------------------------------------------- */
struct vesafb_par {
u32 pseudo_palette[256];
int wc_cookie;
};
static struct fb_var_screeninfo vesafb_defined = {
.activate = FB_ACTIVATE_NOW,
.height = -1,
.width = -1,
.right_margin = 32,
.upper_margin = 16,
.lower_margin = 4,
.vsync_len = 4,
.vmode = FB_VMODE_NONINTERLACED,
};
static struct fb_fix_screeninfo vesafb_fix = {
.id = "VESA VGA",
.type = FB_TYPE_PACKED_PIXELS,
.accel = FB_ACCEL_NONE,
};
static int inverse __read_mostly;
static int mtrr __read_mostly; /* disable mtrr */
static int vram_remap; /* Set amount of memory to be used */
static int vram_total; /* Set total amount of memory */
static int pmi_setpal __read_mostly = 1; /* pmi for palette changes ??? */
static int ypan __read_mostly; /* 0..nothing, 1..ypan, 2..ywrap */
static void (*pmi_start)(void) __read_mostly;
static void (*pmi_pal) (void) __read_mostly;
static int depth __read_mostly;
static int vga_compat __read_mostly;
/* --------------------------------------------------------------------- */
static int vesafb_pan_display(struct fb_var_screeninfo *var,
struct fb_info *info)
{
#ifdef __i386__
int offset;
offset = (var->yoffset * info->fix.line_length + var->xoffset) / 4;
__asm__ __volatile__(
"call *(%%edi)"
: /* no return value */
: "a" (0x4f07), /* EAX */
"b" (0), /* EBX */
"c" (offset), /* ECX */
"d" (offset >> 16), /* EDX */
"D" (&pmi_start)); /* EDI */
#endif
return 0;
}
static int vesa_setpalette(int regno, unsigned red, unsigned green,
unsigned blue)
{
int shift = 16 - depth;
int err = -EINVAL;
/*
* Try VGA registers first...
*/
if (vga_compat) {
outb_p(regno, dac_reg);
outb_p(red >> shift, dac_val);
outb_p(green >> shift, dac_val);
outb_p(blue >> shift, dac_val);
err = 0;
}
#ifdef __i386__
/*
* Fallback to the PMI....
*/
if (err && pmi_setpal) {
struct { u_char blue, green, red, pad; } entry;
entry.red = red >> shift;
entry.green = green >> shift;
entry.blue = blue >> shift;
entry.pad = 0;
__asm__ __volatile__(
"call *(%%esi)"
: /* no return value */
: "a" (0x4f09), /* EAX */
"b" (0), /* EBX */
"c" (1), /* ECX */
"d" (regno), /* EDX */
"D" (&entry), /* EDI */
"S" (&pmi_pal)); /* ESI */
err = 0;
}
#endif
return err;
}
static int vesafb_setcolreg(unsigned regno, unsigned red, unsigned green,
unsigned blue, unsigned transp,
struct fb_info *info)
{
int err = 0;
/*
* Set a single color register. The values supplied are
* already rounded down to the hardware's capabilities
* (according to the entries in the `var' structure). Return
* != 0 for invalid regno.
*/
if (regno >= info->cmap.len)
return 1;
if (info->var.bits_per_pixel == 8)
err = vesa_setpalette(regno,red,green,blue);
else if (regno < 16) {
switch (info->var.bits_per_pixel) {
case 16:
if (info->var.red.offset == 10) {
/* 1:5:5:5 */
((u32*) (info->pseudo_palette))[regno] =
((red & 0xf800) >> 1) |
((green & 0xf800) >> 6) |
((blue & 0xf800) >> 11);
} else {
/* 0:5:6:5 */
((u32*) (info->pseudo_palette))[regno] =
((red & 0xf800) ) |
((green & 0xfc00) >> 5) |
((blue & 0xf800) >> 11);
}
break;
case 24:
case 32:
red >>= 8;
green >>= 8;
blue >>= 8;
((u32 *)(info->pseudo_palette))[regno] =
(red << info->var.red.offset) |
(green << info->var.green.offset) |
(blue << info->var.blue.offset);
break;
}
}
return err;
}
static void vesafb_destroy(struct fb_info *info)
{
struct vesafb_par *par = info->par;
fb_dealloc_cmap(&info->cmap);
video: fbdev: vesafb: use arch_phys_wc_add() This driver uses the same area for MTRR as for the ioremap_wc(), if anything it just uses a smaller size in case MTRR reservation fails. ioremap_wc() API is already used to take advantage of architecture write-combining when available. Convert the driver from using the x86 specific MTRR code to the architecture agnostic arch_phys_wc_add(). arch_phys_wc_add() will avoid MTRR if write-combining is available. There are a few motivations for this: a) Take advantage of PAT when available b) Help bury MTRR code away, MTRR is architecture specific and on x86 its replaced by PAT c) Help with the goal of eventually using _PAGE_CACHE_UC over _PAGE_CACHE_UC_MINUS on x86 on ioremap_nocache() (see commit de33c442e titled "x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()") The conversion done is expressed by the following Coccinelle SmPL patch, it additionally required manual intervention to address all the #ifdery and removal of redundant things which arch_phys_wc_add() already addresses such as verbose message about when MTRR fails and doing nothing when we didn't get an MTRR. @ mtrr_found @ expression index, base, size; @@ -index = mtrr_add(base, size, MTRR_TYPE_WRCOMB, 1); +index = arch_phys_wc_add(base, size); @ mtrr_rm depends on mtrr_found @ expression mtrr_found.index, mtrr_found.base, mtrr_found.size; @@ -mtrr_del(index, base, size); +arch_phys_wc_del(index); @ mtrr_rm_zero_arg depends on mtrr_found @ expression mtrr_found.index; @@ -mtrr_del(index, 0, 0); +arch_phys_wc_del(index); @ mtrr_rm_fb_info depends on mtrr_found @ struct fb_info *info; expression mtrr_found.index; @@ -mtrr_del(index, info->fix.smem_start, info->fix.smem_len); +arch_phys_wc_del(index); @ ioremap_replace_nocache depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap_nocache(base, size); +info->screen_base = ioremap_wc(base, size); @ ioremap_replace_default depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap(base, size); +info->screen_base = ioremap_wc(base, size); Generated-by: Coccinelle SmPL Cc: Toshi Kani <toshi.kani@hp.com> Cc: Suresh Siddha <sbsiddha@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Juergen Gross <jgross@suse.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Dave Airlie <airlied@redhat.com> Cc: Antonino Daplas <adaplas@gmail.com> Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Jingoo Han <jg1.han@samsung.com> Cc: Wolfram Sang <wsa@the-dreams.de> Cc: Tomi Valkeinen <tomi.valkeinen@ti.com> Cc: linux-fbdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Luis R. Rodriguez <mcgrof@suse.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
2015-06-04 16:44:48 +00:00
arch_phys_wc_del(par->wc_cookie);
if (info->screen_base)
iounmap(info->screen_base);
release_mem_region(info->apertures->ranges[0].base, info->apertures->ranges[0].size);
}
static struct fb_ops vesafb_ops = {
.owner = THIS_MODULE,
.fb_destroy = vesafb_destroy,
.fb_setcolreg = vesafb_setcolreg,
.fb_pan_display = vesafb_pan_display,
.fb_fillrect = cfb_fillrect,
.fb_copyarea = cfb_copyarea,
.fb_imageblit = cfb_imageblit,
};
static int vesafb_setup(char *options)
{
char *this_opt;
if (!options || !*options)
return 0;
while ((this_opt = strsep(&options, ",")) != NULL) {
if (!*this_opt) continue;
if (! strcmp(this_opt, "inverse"))
inverse=1;
else if (! strcmp(this_opt, "redraw"))
ypan=0;
else if (! strcmp(this_opt, "ypan"))
ypan=1;
else if (! strcmp(this_opt, "ywrap"))
ypan=2;
else if (! strcmp(this_opt, "vgapal"))
pmi_setpal=0;
else if (! strcmp(this_opt, "pmipal"))
pmi_setpal=1;
else if (! strncmp(this_opt, "mtrr:", 5))
mtrr = simple_strtoul(this_opt+5, NULL, 0);
else if (! strcmp(this_opt, "nomtrr"))
mtrr=0;
else if (! strncmp(this_opt, "vtotal:", 7))
vram_total = simple_strtoul(this_opt+7, NULL, 0);
else if (! strncmp(this_opt, "vremap:", 7))
vram_remap = simple_strtoul(this_opt+7, NULL, 0);
}
return 0;
}
static int vesafb_probe(struct platform_device *dev)
{
struct fb_info *info;
struct vesafb_par *par;
int i, err;
unsigned int size_vmode;
unsigned int size_remap;
unsigned int size_total;
char *option = NULL;
/* ignore error return of fb_get_options */
fb_get_options("vesafb", &option);
vesafb_setup(option);
if (screen_info.orig_video_isVGA != VIDEO_TYPE_VLFB)
return -ENODEV;
vga_compat = (screen_info.capabilities & 2) ? 0 : 1;
vesafb_fix.smem_start = screen_info.lfb_base;
vesafb_defined.bits_per_pixel = screen_info.lfb_depth;
if (15 == vesafb_defined.bits_per_pixel)
vesafb_defined.bits_per_pixel = 16;
vesafb_defined.xres = screen_info.lfb_width;
vesafb_defined.yres = screen_info.lfb_height;
vesafb_fix.line_length = screen_info.lfb_linelength;
vesafb_fix.visual = (vesafb_defined.bits_per_pixel == 8) ?
FB_VISUAL_PSEUDOCOLOR : FB_VISUAL_TRUECOLOR;
/* size_vmode -- that is the amount of memory needed for the
* used video mode, i.e. the minimum amount of
* memory we need. */
size_vmode = vesafb_defined.yres * vesafb_fix.line_length;
/* size_total -- all video memory we have. Used for mtrr
* entries, resource allocation and bounds
* checking. */
size_total = screen_info.lfb_size * 65536;
if (vram_total)
size_total = vram_total * 1024 * 1024;
if (size_total < size_vmode)
size_total = size_vmode;
/* size_remap -- the amount of video memory we are going to
* use for vesafb. With modern cards it is no
* option to simply use size_total as that
* wastes plenty of kernel address space. */
size_remap = size_vmode * 2;
if (vram_remap)
size_remap = vram_remap * 1024 * 1024;
if (size_remap < size_vmode)
size_remap = size_vmode;
if (size_remap > size_total)
size_remap = size_total;
vesafb_fix.smem_len = size_remap;
#ifndef __i386__
screen_info.vesapm_seg = 0;
#endif
if (!request_mem_region(vesafb_fix.smem_start, size_total, "vesafb")) {
printk(KERN_WARNING
"vesafb: cannot reserve video memory at 0x%lx\n",
vesafb_fix.smem_start);
/* We cannot make this fatal. Sometimes this comes from magic
spaces our resource handlers simply don't know about */
}
info = framebuffer_alloc(sizeof(struct vesafb_par), &dev->dev);
if (!info) {
release_mem_region(vesafb_fix.smem_start, size_total);
return -ENOMEM;
}
platform_set_drvdata(dev, info);
par = info->par;
info->pseudo_palette = par->pseudo_palette;
/* set vesafb aperture size for generic probing */
info->apertures = alloc_apertures(1);
if (!info->apertures) {
err = -ENOMEM;
goto err;
}
info->apertures->ranges[0].base = screen_info.lfb_base;
info->apertures->ranges[0].size = size_total;
printk(KERN_INFO "vesafb: mode is %dx%dx%d, linelength=%d, pages=%d\n",
vesafb_defined.xres, vesafb_defined.yres, vesafb_defined.bits_per_pixel, vesafb_fix.line_length, screen_info.pages);
if (screen_info.vesapm_seg) {
printk(KERN_INFO "vesafb: protected mode interface info at %04x:%04x\n",
screen_info.vesapm_seg,screen_info.vesapm_off);
}
if (screen_info.vesapm_seg < 0xc000)
ypan = pmi_setpal = 0; /* not available or some DOS TSR ... */
if (ypan || pmi_setpal) {
unsigned short *pmi_base;
pmi_base = (unsigned short*)phys_to_virt(((unsigned long)screen_info.vesapm_seg << 4) + screen_info.vesapm_off);
pmi_start = (void*)((char*)pmi_base + pmi_base[1]);
pmi_pal = (void*)((char*)pmi_base + pmi_base[2]);
printk(KERN_INFO "vesafb: pmi: set display start = %p, set palette = %p\n",pmi_start,pmi_pal);
if (pmi_base[3]) {
printk(KERN_INFO "vesafb: pmi: ports = ");
for (i = pmi_base[3]/2; pmi_base[i] != 0xffff; i++)
printk("%x ",pmi_base[i]);
printk("\n");
if (pmi_base[i] != 0xffff) {
/*
* memory areas not supported (yet?)
*
* Rules are: we have to set up a descriptor for the requested
* memory area and pass it in the ES register to the BIOS function.
*/
printk(KERN_INFO "vesafb: can't handle memory requests, pmi disabled\n");
ypan = pmi_setpal = 0;
}
}
}
if (vesafb_defined.bits_per_pixel == 8 && !pmi_setpal && !vga_compat) {
printk(KERN_WARNING "vesafb: hardware palette is unchangeable,\n"
" colors may be incorrect\n");
vesafb_fix.visual = FB_VISUAL_STATIC_PSEUDOCOLOR;
}
vesafb_defined.xres_virtual = vesafb_defined.xres;
vesafb_defined.yres_virtual = vesafb_fix.smem_len / vesafb_fix.line_length;
if (ypan && vesafb_defined.yres_virtual > vesafb_defined.yres) {
printk(KERN_INFO "vesafb: scrolling: %s using protected mode interface, yres_virtual=%d\n",
(ypan > 1) ? "ywrap" : "ypan",vesafb_defined.yres_virtual);
} else {
printk(KERN_INFO "vesafb: scrolling: redraw\n");
vesafb_defined.yres_virtual = vesafb_defined.yres;
ypan = 0;
}
/* some dummy values for timing to make fbset happy */
vesafb_defined.pixclock = 10000000 / vesafb_defined.xres * 1000 / vesafb_defined.yres;
vesafb_defined.left_margin = (vesafb_defined.xres / 8) & 0xf8;
vesafb_defined.hsync_len = (vesafb_defined.xres / 8) & 0xf8;
vesafb_defined.red.offset = screen_info.red_pos;
vesafb_defined.red.length = screen_info.red_size;
vesafb_defined.green.offset = screen_info.green_pos;
vesafb_defined.green.length = screen_info.green_size;
vesafb_defined.blue.offset = screen_info.blue_pos;
vesafb_defined.blue.length = screen_info.blue_size;
vesafb_defined.transp.offset = screen_info.rsvd_pos;
vesafb_defined.transp.length = screen_info.rsvd_size;
if (vesafb_defined.bits_per_pixel <= 8) {
depth = vesafb_defined.green.length;
vesafb_defined.red.length =
vesafb_defined.green.length =
vesafb_defined.blue.length =
vesafb_defined.bits_per_pixel;
}
printk(KERN_INFO "vesafb: %s: "
"size=%d:%d:%d:%d, shift=%d:%d:%d:%d\n",
(vesafb_defined.bits_per_pixel > 8) ?
"Truecolor" : (vga_compat || pmi_setpal) ?
"Pseudocolor" : "Static Pseudocolor",
screen_info.rsvd_size,
screen_info.red_size,
screen_info.green_size,
screen_info.blue_size,
screen_info.rsvd_pos,
screen_info.red_pos,
screen_info.green_pos,
screen_info.blue_pos);
vesafb_fix.ypanstep = ypan ? 1 : 0;
vesafb_fix.ywrapstep = (ypan>1) ? 1 : 0;
/* request failure does not faze us, as vgacon probably has this
* region already (FIXME) */
request_region(0x3c0, 32, "vesafb");
if (mtrr == 3) {
unsigned int temp_size = size_total;
/* Find the largest power-of-two */
temp_size = roundup_pow_of_two(temp_size);
/* Try and find a power of two to add */
do {
video: fbdev: vesafb: use arch_phys_wc_add() This driver uses the same area for MTRR as for the ioremap_wc(), if anything it just uses a smaller size in case MTRR reservation fails. ioremap_wc() API is already used to take advantage of architecture write-combining when available. Convert the driver from using the x86 specific MTRR code to the architecture agnostic arch_phys_wc_add(). arch_phys_wc_add() will avoid MTRR if write-combining is available. There are a few motivations for this: a) Take advantage of PAT when available b) Help bury MTRR code away, MTRR is architecture specific and on x86 its replaced by PAT c) Help with the goal of eventually using _PAGE_CACHE_UC over _PAGE_CACHE_UC_MINUS on x86 on ioremap_nocache() (see commit de33c442e titled "x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()") The conversion done is expressed by the following Coccinelle SmPL patch, it additionally required manual intervention to address all the #ifdery and removal of redundant things which arch_phys_wc_add() already addresses such as verbose message about when MTRR fails and doing nothing when we didn't get an MTRR. @ mtrr_found @ expression index, base, size; @@ -index = mtrr_add(base, size, MTRR_TYPE_WRCOMB, 1); +index = arch_phys_wc_add(base, size); @ mtrr_rm depends on mtrr_found @ expression mtrr_found.index, mtrr_found.base, mtrr_found.size; @@ -mtrr_del(index, base, size); +arch_phys_wc_del(index); @ mtrr_rm_zero_arg depends on mtrr_found @ expression mtrr_found.index; @@ -mtrr_del(index, 0, 0); +arch_phys_wc_del(index); @ mtrr_rm_fb_info depends on mtrr_found @ struct fb_info *info; expression mtrr_found.index; @@ -mtrr_del(index, info->fix.smem_start, info->fix.smem_len); +arch_phys_wc_del(index); @ ioremap_replace_nocache depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap_nocache(base, size); +info->screen_base = ioremap_wc(base, size); @ ioremap_replace_default depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap(base, size); +info->screen_base = ioremap_wc(base, size); Generated-by: Coccinelle SmPL Cc: Toshi Kani <toshi.kani@hp.com> Cc: Suresh Siddha <sbsiddha@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Juergen Gross <jgross@suse.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Dave Airlie <airlied@redhat.com> Cc: Antonino Daplas <adaplas@gmail.com> Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Jingoo Han <jg1.han@samsung.com> Cc: Wolfram Sang <wsa@the-dreams.de> Cc: Tomi Valkeinen <tomi.valkeinen@ti.com> Cc: linux-fbdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Luis R. Rodriguez <mcgrof@suse.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
2015-06-04 16:44:48 +00:00
par->wc_cookie =
arch_phys_wc_add(vesafb_fix.smem_start,
temp_size);
temp_size >>= 1;
video: fbdev: vesafb: use arch_phys_wc_add() This driver uses the same area for MTRR as for the ioremap_wc(), if anything it just uses a smaller size in case MTRR reservation fails. ioremap_wc() API is already used to take advantage of architecture write-combining when available. Convert the driver from using the x86 specific MTRR code to the architecture agnostic arch_phys_wc_add(). arch_phys_wc_add() will avoid MTRR if write-combining is available. There are a few motivations for this: a) Take advantage of PAT when available b) Help bury MTRR code away, MTRR is architecture specific and on x86 its replaced by PAT c) Help with the goal of eventually using _PAGE_CACHE_UC over _PAGE_CACHE_UC_MINUS on x86 on ioremap_nocache() (see commit de33c442e titled "x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()") The conversion done is expressed by the following Coccinelle SmPL patch, it additionally required manual intervention to address all the #ifdery and removal of redundant things which arch_phys_wc_add() already addresses such as verbose message about when MTRR fails and doing nothing when we didn't get an MTRR. @ mtrr_found @ expression index, base, size; @@ -index = mtrr_add(base, size, MTRR_TYPE_WRCOMB, 1); +index = arch_phys_wc_add(base, size); @ mtrr_rm depends on mtrr_found @ expression mtrr_found.index, mtrr_found.base, mtrr_found.size; @@ -mtrr_del(index, base, size); +arch_phys_wc_del(index); @ mtrr_rm_zero_arg depends on mtrr_found @ expression mtrr_found.index; @@ -mtrr_del(index, 0, 0); +arch_phys_wc_del(index); @ mtrr_rm_fb_info depends on mtrr_found @ struct fb_info *info; expression mtrr_found.index; @@ -mtrr_del(index, info->fix.smem_start, info->fix.smem_len); +arch_phys_wc_del(index); @ ioremap_replace_nocache depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap_nocache(base, size); +info->screen_base = ioremap_wc(base, size); @ ioremap_replace_default depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap(base, size); +info->screen_base = ioremap_wc(base, size); Generated-by: Coccinelle SmPL Cc: Toshi Kani <toshi.kani@hp.com> Cc: Suresh Siddha <sbsiddha@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Juergen Gross <jgross@suse.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Dave Airlie <airlied@redhat.com> Cc: Antonino Daplas <adaplas@gmail.com> Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Jingoo Han <jg1.han@samsung.com> Cc: Wolfram Sang <wsa@the-dreams.de> Cc: Tomi Valkeinen <tomi.valkeinen@ti.com> Cc: linux-fbdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Luis R. Rodriguez <mcgrof@suse.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
2015-06-04 16:44:48 +00:00
} while (temp_size >= PAGE_SIZE && par->wc_cookie < 0);
uvesafb,vesafb: create WC or WB PAT-entries with an PAT-enabled kernel, when using uvesafb or vesafb, these drivers will create uncached-minus PAT entries for the framebuffer memory because they use ioremap() (not the *_cache or *_wc variants). When the framebuffer memory intersects with the video RAM used by Xorg, the complete video RAM will be mapped uncached-minus what results in a serve performance penalty. Here are the correct MTRR entries created by uvesafb: schlicht@netbook:~$ cat /proc/mtrr reg00: base=0x000000000 ( 0MB), size= 2048MB, count=1: write-back reg01: base=0x06ff00000 ( 1791MB), size= 1MB, count=1: uncachable reg02: base=0x070000000 ( 1792MB), size= 256MB, count=1: uncachable reg03: base=0x0d0000000 ( 3328MB), size= 16MB, count=1: write-combining And here are the problematic PAT entries: schlicht@netbook:~$ sudo cat /sys/kernel/debug/x86/pat_memtype_list PAT memtype list: write-back @ 0x0-0x1000 uncached-minus @ 0x6fedd000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0xd0000000-0xe0000000 <-- created by xserver-xorg uncached-minus @ 0xd0000000-0xd1194000 <-- created by uvesafb uncached-minus @ 0xf4000000-0xf4009000 uncached-minus @ 0xf4200000-0xf4400000 uncached-minus @ 0xf5000000-0xf5010000 uncached-minus @ 0xf5100000-0xf5104000 uncached-minus @ 0xf5400000-0xf5404000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xfed00000-0xfed01000 Therefore I created the attached patch for uvesafb which uses ioremap_wc() to create the correct PAT entries, as shown below: schlicht@netbook:~$ sudo cat /sys/kernel/debug/x86/pat_memtype_list PAT memtype list: write-back @ 0x0-0x1000 uncached-minus @ 0x6fedd000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 write-combining @ 0xd0000000-0xe0000000 write-combining @ 0xd0000000-0xd1194000 uncached-minus @ 0xf4000000-0xf4009000 uncached-minus @ 0xf4200000-0xf4400000 uncached-minus @ 0xf5000000-0xf5010000 uncached-minus @ 0xf5100000-0xf5104000 uncached-minus @ 0xf5400000-0xf5404000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xfed00000-0xfed01000 This results in a performance gain, objectively measurable with e.g. x11perf -comppixwin10 -comppixwin100 -comppixwin500: 1: x11perf_xaa.log 2: x11perf_xaa_patched.log 1 2 Operation -------- ---------------- ----------------- 124000.0 202000.0 ( 1.63) Composite 10x10 from pixmap to window 3340.0 24400.0 ( 7.31) Composite 100x100 from pixmap to window 131.0 1150.0 ( 8.78) Composite 500x500 from pixmap to window You can see the serve performance gain when composing larger pixmaps to window. The patches replace the ioremap() function with the variant matching the mtrr- parameter. To create "write-back" PAT entries, the ioremap_cache() function must be called after creating the MTRR entries, and the ioremap_cache() region must completely fit into the MTRR region, this is why the MTRR region size is now rounded up to the next power-of-two. Signed-off-by: Thomas Schlichter <thomas.schlichter@web.de> Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2010-11-27 13:17:55 +00:00
info->screen_base = ioremap_wc(vesafb_fix.smem_start, vesafb_fix.smem_len);
} else {
if (mtrr && mtrr != 3)
WARN_ONCE(1, "Only MTRR_TYPE_WRCOMB (3) make sense\n");
uvesafb,vesafb: create WC or WB PAT-entries with an PAT-enabled kernel, when using uvesafb or vesafb, these drivers will create uncached-minus PAT entries for the framebuffer memory because they use ioremap() (not the *_cache or *_wc variants). When the framebuffer memory intersects with the video RAM used by Xorg, the complete video RAM will be mapped uncached-minus what results in a serve performance penalty. Here are the correct MTRR entries created by uvesafb: schlicht@netbook:~$ cat /proc/mtrr reg00: base=0x000000000 ( 0MB), size= 2048MB, count=1: write-back reg01: base=0x06ff00000 ( 1791MB), size= 1MB, count=1: uncachable reg02: base=0x070000000 ( 1792MB), size= 256MB, count=1: uncachable reg03: base=0x0d0000000 ( 3328MB), size= 16MB, count=1: write-combining And here are the problematic PAT entries: schlicht@netbook:~$ sudo cat /sys/kernel/debug/x86/pat_memtype_list PAT memtype list: write-back @ 0x0-0x1000 uncached-minus @ 0x6fedd000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0xd0000000-0xe0000000 <-- created by xserver-xorg uncached-minus @ 0xd0000000-0xd1194000 <-- created by uvesafb uncached-minus @ 0xf4000000-0xf4009000 uncached-minus @ 0xf4200000-0xf4400000 uncached-minus @ 0xf5000000-0xf5010000 uncached-minus @ 0xf5100000-0xf5104000 uncached-minus @ 0xf5400000-0xf5404000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xfed00000-0xfed01000 Therefore I created the attached patch for uvesafb which uses ioremap_wc() to create the correct PAT entries, as shown below: schlicht@netbook:~$ sudo cat /sys/kernel/debug/x86/pat_memtype_list PAT memtype list: write-back @ 0x0-0x1000 uncached-minus @ 0x6fedd000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 write-combining @ 0xd0000000-0xe0000000 write-combining @ 0xd0000000-0xd1194000 uncached-minus @ 0xf4000000-0xf4009000 uncached-minus @ 0xf4200000-0xf4400000 uncached-minus @ 0xf5000000-0xf5010000 uncached-minus @ 0xf5100000-0xf5104000 uncached-minus @ 0xf5400000-0xf5404000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xfed00000-0xfed01000 This results in a performance gain, objectively measurable with e.g. x11perf -comppixwin10 -comppixwin100 -comppixwin500: 1: x11perf_xaa.log 2: x11perf_xaa_patched.log 1 2 Operation -------- ---------------- ----------------- 124000.0 202000.0 ( 1.63) Composite 10x10 from pixmap to window 3340.0 24400.0 ( 7.31) Composite 100x100 from pixmap to window 131.0 1150.0 ( 8.78) Composite 500x500 from pixmap to window You can see the serve performance gain when composing larger pixmaps to window. The patches replace the ioremap() function with the variant matching the mtrr- parameter. To create "write-back" PAT entries, the ioremap_cache() function must be called after creating the MTRR entries, and the ioremap_cache() region must completely fit into the MTRR region, this is why the MTRR region size is now rounded up to the next power-of-two. Signed-off-by: Thomas Schlichter <thomas.schlichter@web.de> Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2010-11-27 13:17:55 +00:00
info->screen_base = ioremap(vesafb_fix.smem_start, vesafb_fix.smem_len);
}
uvesafb,vesafb: create WC or WB PAT-entries with an PAT-enabled kernel, when using uvesafb or vesafb, these drivers will create uncached-minus PAT entries for the framebuffer memory because they use ioremap() (not the *_cache or *_wc variants). When the framebuffer memory intersects with the video RAM used by Xorg, the complete video RAM will be mapped uncached-minus what results in a serve performance penalty. Here are the correct MTRR entries created by uvesafb: schlicht@netbook:~$ cat /proc/mtrr reg00: base=0x000000000 ( 0MB), size= 2048MB, count=1: write-back reg01: base=0x06ff00000 ( 1791MB), size= 1MB, count=1: uncachable reg02: base=0x070000000 ( 1792MB), size= 256MB, count=1: uncachable reg03: base=0x0d0000000 ( 3328MB), size= 16MB, count=1: write-combining And here are the problematic PAT entries: schlicht@netbook:~$ sudo cat /sys/kernel/debug/x86/pat_memtype_list PAT memtype list: write-back @ 0x0-0x1000 uncached-minus @ 0x6fedd000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0xd0000000-0xe0000000 <-- created by xserver-xorg uncached-minus @ 0xd0000000-0xd1194000 <-- created by uvesafb uncached-minus @ 0xf4000000-0xf4009000 uncached-minus @ 0xf4200000-0xf4400000 uncached-minus @ 0xf5000000-0xf5010000 uncached-minus @ 0xf5100000-0xf5104000 uncached-minus @ 0xf5400000-0xf5404000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xfed00000-0xfed01000 Therefore I created the attached patch for uvesafb which uses ioremap_wc() to create the correct PAT entries, as shown below: schlicht@netbook:~$ sudo cat /sys/kernel/debug/x86/pat_memtype_list PAT memtype list: write-back @ 0x0-0x1000 uncached-minus @ 0x6fedd000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee2000-0x6fee3000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 uncached-minus @ 0x6fee3000-0x6fee4000 write-combining @ 0xd0000000-0xe0000000 write-combining @ 0xd0000000-0xd1194000 uncached-minus @ 0xf4000000-0xf4009000 uncached-minus @ 0xf4200000-0xf4400000 uncached-minus @ 0xf5000000-0xf5010000 uncached-minus @ 0xf5100000-0xf5104000 uncached-minus @ 0xf5400000-0xf5404000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xf5404000-0xf5405000 uncached-minus @ 0xfed00000-0xfed01000 This results in a performance gain, objectively measurable with e.g. x11perf -comppixwin10 -comppixwin100 -comppixwin500: 1: x11perf_xaa.log 2: x11perf_xaa_patched.log 1 2 Operation -------- ---------------- ----------------- 124000.0 202000.0 ( 1.63) Composite 10x10 from pixmap to window 3340.0 24400.0 ( 7.31) Composite 100x100 from pixmap to window 131.0 1150.0 ( 8.78) Composite 500x500 from pixmap to window You can see the serve performance gain when composing larger pixmaps to window. The patches replace the ioremap() function with the variant matching the mtrr- parameter. To create "write-back" PAT entries, the ioremap_cache() function must be called after creating the MTRR entries, and the ioremap_cache() region must completely fit into the MTRR region, this is why the MTRR region size is now rounded up to the next power-of-two. Signed-off-by: Thomas Schlichter <thomas.schlichter@web.de> Signed-off-by: Paul Mundt <lethal@linux-sh.org>
2010-11-27 13:17:55 +00:00
if (!info->screen_base) {
printk(KERN_ERR
"vesafb: abort, cannot ioremap video memory 0x%x @ 0x%lx\n",
vesafb_fix.smem_len, vesafb_fix.smem_start);
err = -EIO;
goto err;
}
printk(KERN_INFO "vesafb: framebuffer at 0x%lx, mapped to 0x%p, "
"using %dk, total %dk\n",
vesafb_fix.smem_start, info->screen_base,
size_remap/1024, size_total/1024);
info->fbops = &vesafb_ops;
info->var = vesafb_defined;
info->fix = vesafb_fix;
info->flags = FBINFO_FLAG_DEFAULT | FBINFO_MISC_FIRMWARE |
(ypan ? FBINFO_HWACCEL_YPAN : 0);
if (!ypan)
info->fbops->fb_pan_display = NULL;
if (fb_alloc_cmap(&info->cmap, 256, 0) < 0) {
err = -ENOMEM;
goto err;
}
if (register_framebuffer(info)<0) {
err = -EINVAL;
fb_dealloc_cmap(&info->cmap);
goto err;
}
fb_info(info, "%s frame buffer device\n", info->fix.id);
return 0;
err:
video: fbdev: vesafb: use arch_phys_wc_add() This driver uses the same area for MTRR as for the ioremap_wc(), if anything it just uses a smaller size in case MTRR reservation fails. ioremap_wc() API is already used to take advantage of architecture write-combining when available. Convert the driver from using the x86 specific MTRR code to the architecture agnostic arch_phys_wc_add(). arch_phys_wc_add() will avoid MTRR if write-combining is available. There are a few motivations for this: a) Take advantage of PAT when available b) Help bury MTRR code away, MTRR is architecture specific and on x86 its replaced by PAT c) Help with the goal of eventually using _PAGE_CACHE_UC over _PAGE_CACHE_UC_MINUS on x86 on ioremap_nocache() (see commit de33c442e titled "x86 PAT: fix performance drop for glx, use UC minus for ioremap(), ioremap_nocache() and pci_mmap_page_range()") The conversion done is expressed by the following Coccinelle SmPL patch, it additionally required manual intervention to address all the #ifdery and removal of redundant things which arch_phys_wc_add() already addresses such as verbose message about when MTRR fails and doing nothing when we didn't get an MTRR. @ mtrr_found @ expression index, base, size; @@ -index = mtrr_add(base, size, MTRR_TYPE_WRCOMB, 1); +index = arch_phys_wc_add(base, size); @ mtrr_rm depends on mtrr_found @ expression mtrr_found.index, mtrr_found.base, mtrr_found.size; @@ -mtrr_del(index, base, size); +arch_phys_wc_del(index); @ mtrr_rm_zero_arg depends on mtrr_found @ expression mtrr_found.index; @@ -mtrr_del(index, 0, 0); +arch_phys_wc_del(index); @ mtrr_rm_fb_info depends on mtrr_found @ struct fb_info *info; expression mtrr_found.index; @@ -mtrr_del(index, info->fix.smem_start, info->fix.smem_len); +arch_phys_wc_del(index); @ ioremap_replace_nocache depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap_nocache(base, size); +info->screen_base = ioremap_wc(base, size); @ ioremap_replace_default depends on mtrr_found @ struct fb_info *info; expression base, size; @@ -info->screen_base = ioremap(base, size); +info->screen_base = ioremap_wc(base, size); Generated-by: Coccinelle SmPL Cc: Toshi Kani <toshi.kani@hp.com> Cc: Suresh Siddha <sbsiddha@gmail.com> Cc: Ingo Molnar <mingo@elte.hu> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Juergen Gross <jgross@suse.com> Cc: Daniel Vetter <daniel.vetter@ffwll.ch> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Dave Airlie <airlied@redhat.com> Cc: Antonino Daplas <adaplas@gmail.com> Cc: Jean-Christophe Plagniol-Villard <plagnioj@jcrosoft.com> Cc: Rob Clark <robdclark@gmail.com> Cc: Jingoo Han <jg1.han@samsung.com> Cc: Wolfram Sang <wsa@the-dreams.de> Cc: Tomi Valkeinen <tomi.valkeinen@ti.com> Cc: linux-fbdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Signed-off-by: Luis R. Rodriguez <mcgrof@suse.com> Signed-off-by: Tomi Valkeinen <tomi.valkeinen@ti.com>
2015-06-04 16:44:48 +00:00
arch_phys_wc_del(par->wc_cookie);
if (info->screen_base)
iounmap(info->screen_base);
framebuffer_release(info);
release_mem_region(vesafb_fix.smem_start, size_total);
return err;
}
static int vesafb_remove(struct platform_device *pdev)
{
struct fb_info *info = platform_get_drvdata(pdev);
unregister_framebuffer(info);
framebuffer_release(info);
return 0;
}
static struct platform_driver vesafb_driver = {
.driver = {
.name = "vesa-framebuffer",
},
.probe = vesafb_probe,
.remove = vesafb_remove,
};
module_platform_driver(vesafb_driver);
MODULE_LICENSE("GPL");