Re: [Suspend-devel] "Update" code to vbetool-0.6

[Stefan Seyfried]

On Tue, Jul 18, 2006 at 12:52:44PM +0200, Stefan Seyfried wrote:
> Hi,
> 
> i have the following in my queue. This updates the code to the vbetool 0.6
> codebase. There is nothing in there which should actually make any difference
> to s2ram, but it might reduce the "you don't even use the latest vbetool
> code!!!1! no wonder this does not work!!!1!" mails i get :-)

ok, i actually fscked it up, this is what i forgot in my merge.
Again, these are not actually used inside s2ram so it should not matter.
If it does make a difference in code size, we could probably wrap them
inside #ifdef S2RAM or something like that.

This goes in on top of the (already committed, sorry) "upgrade to
vbetool-0.6" patch.

Index: vbetool/vbetool.c
===================================================================
RCS file: /cvsroot/suspend/suspend/vbetool/vbetool.c,v
retrieving revision 1.5
diff -u -p -r1.5 vbetool.c
--- vbetool/vbetool.c   19 Jul 2006 07:38:07 -0000      1.5
+++ vbetool/vbetool.c   19 Jul 2006 08:42:01 -0000
@@ -379,6 +379,68 @@ int do_set_mode (int mode, int vga) {
        return 0;
 }
 
+int do_get_panel_brightness() {
+       reg_frame regs;
+       int error;
+
+       memset(&regs, 0, sizeof(regs));
+
+       error = do_vbe_service(0x4f11, 0x05, &regs);
+
+       if (error<0) {
+               return error;
+       }
+
+       printf("%d\n",regs.ecx);
+
+       return 0;
+}
+
+int do_invert_panel() {
+       reg_frame regs;
+       int error;
+
+       memset(&regs, 0, sizeof(regs));
+
+       error = do_vbe_service(0x4f11, 0x02, &regs);
+
+       if (error<0) {
+               return error;
+       }
+
+       if ((regs.ebx & 0xff) == 0)
+               regs.ecx = 3;
+       else
+               regs.ecx = 0;
+
+       error = do_vbe_service(0x4f11, 0x0102, &regs);
+
+       if (error<0) {
+               return error;
+       }
+
+       return 0;
+}
+
+int do_set_panel_brightness(int brightness) {
+       reg_frame regs;
+       int error;
+
+       memset(&regs, 0, sizeof(regs));
+
+       regs.ecx = brightness;
+
+       error = do_vbe_service(0x4f11, 0x0105, &regs);
+
+       if (error<0) {
+               return error;
+       }
+
+       printf("%d\n",regs.ecx);
+
+       return 0;
+}
+
 int do_get_mode() {
        reg_frame regs;
        int error;
@@ -430,3 +492,65 @@ int disable_vga() {
        outb(~0x01 & inb(0x102),  0x102);
        return 0;
 }
+
+/* Based on xserver-xorg-driver-i810/src/i830_driver.c */
+struct panel_id {
+  int hsize:16, vsize:16;
+  int fptype:16;
+  int redbpp:8, greenbpp:8, bluebpp:8, reservedbpp:8;
+  int rsvdoffscrnmemsize:32, rsvdoffscrnmemptr:32;
+  char reserved[14];
+} __attribute__((packed));
+
+int do_get_panel_id(int just_dimensions)
+{
+  reg_frame r = {
+    .eax = 0x4f11,
+    .ebx = 0x0001
+  };
+  struct panel_id *id = LRMI_alloc_real(sizeof(struct panel_id));
+  r.es = (unsigned short)(((int)(id-LRMI_base_addr()) >> 4) & 0xffff);
+  r.edi = (unsigned long)(id-LRMI_base_addr()) & 0xf;
+
+  if(sizeof(struct panel_id) != 32)
+    return fprintf(stderr, "oops: panel_id, sizeof struct panel_id != 32, it's 
%d...\n", sizeof(struct panel_id)), 7;
+
+  if(real_mode_int(0x10, &r))
+    return fprintf(stderr, "Can't get panel id (vm86 failure)\n"), 8;
+
+  if((r.eax & 0xff) != 0x4f)
+    return fprintf(stderr, "Panel id function not supported\n"), 9;
+
+  if(r.eax & 0xff00)
+    {
+      if((r.eax & 0xff00) == 0x100)
+       fprintf(stderr, "Panel id read failed\n");
+      else
+       fprintf(stderr, "Panel id function not successful\n");
+      return 10;
+    }
+
+  if(!just_dimensions)
+    printf("size:\t%d %d\n"
+          "type:\t%d\n"
+          "bpp:\t%d %d %d %d\n",
+          id->hsize, id->vsize,
+          id->fptype,
+          id->redbpp, id->greenbpp, id->bluebpp, id->reservedbpp);
+  else
+    printf("%dx%d\n", id->hsize, id->vsize);
+
+#if 0
+
+  /* Don't have a use for these and they don't seem to be documented.
+   * 320 appears to be 320kB of mapped memory and the following
+   * pointer is 0x1ffb8000 which is kernel mapping + 0xb8000 offset.
+   */
+  printf("ram:\t%dkB\n"
+        "offset:\t%p\n",
+        id->rsvdoffscrnmemsize,
+        (void *)id->rsvdoffscrnmemptr);
+#endif
+
+  return 0;
+}

-- 
Stefan Seyfried                     | "Please, just tell people
QA / R&D Team Mobile Devices        |               to use KDE."
SUSE LINUX Products GmbH, Nürnberg  |          -- Linus Torvalds

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys -- and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Suspend-devel mailing list
Suspend-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/suspend-devel