HD 7xxx-Can we eliminate the Sleep Trick?
Started by eep357, Jan 18 2013 09:42 PM
7970 7xxx sleep trick
160 replies to this topic
#81
Posted 14 February 2013 - 03:56 AM
ok tested and it sleep fine , but never wake up.. until i press the power button.
#82
Posted 14 February 2013 - 06:51 AM
#83
Posted 14 February 2013 - 10:29 AM
systemsetup -setusingnetworktime off | systemsetup -settime 12:00:00 | pmset sleepnowThis turns off internet updated time and sets the clock to 12:00:00 right before putting the computer to sleep, then the scheduled wake event will wake it back up a few seconds later. So then my next step is to run another script that runs on wake which will set back internet updated time using
#!/bin/bash ntpdate -u time.apple.comto do this I'm using (and still fiddling with) an utility called SleepWatcher (launch daemon) than can run scripts on sleep or wake. The tricky part is the above command requires sudo, and I don't wan't to have to enter my password to have the time corrected when it wakes up. It's possible to do this without having to enter password and there are a few ways, but one must be very careful not to open a security hole in the process. For this reason I wouldn't want to give specific directions and I'm honestly still fiddling with the best/secure method of doing this. The automated sleep/wake part works a treat! But if I had to interact in anyway to restore my system time after each wake, it would take any gained convenience out.
suppose
systemsetup -setusingnetworktime onmight work instead of ntpdate and save that whole headache, will have to give that a go
#84
Posted 14 February 2013 - 02:50 PM
BTW, got it all working great!, as outlined above and using
systemsetup -setusingnetworktime onon wake up. Only caveat is wake time will be rounded to nearest minute by system, so set wake time to 00:01 (can just do this in system preferences) and system time change to 00:00:50 giving it a 10 second sleep. If it takes a while to sleep and shut everything down and it misses it's wake event it won't wake, so finding the right delay is key. Mine takes a while (actually set mine for like 30 secs) because got over 10 hard drives that gotta do their thing before it fully sleeps. Once it wakes up, time is set to normal and I never even see the time change except maybe for a second right when the screen wakes up. Not as fast as EFI/UEFI, but I don't reboot that often, and at least I don't have to babysit it when I do.
#85
Posted 17 February 2013 - 07:36 PM
In case your vendor doesn't provide any hybrid vbios, it is possible to create your own.
I used my XFX legacy vbios and added the EFI part from a similar MSI card.
The XFX legacy vbios got following structure:
1) Legacy part from 0x00000 to 0x0FFFF
0x00000-0x00001: Signature = 0x55 0xAA
0x00002: Length = 0x80 -> 0x80 * 0x200 (block length) = 0x10000
2) Unused space from 0x10000 to 0x1FFFF
The MSI hybrid vbios - on the other hand - looks like this:
1) Legacy part from 0x00000 to 0x0FFFF
0x00000-0x00001: Signature = 0x55 0xAA
0x00002: Length = 0x80 -> 0x80 * 0x200 = 0x10000
2) EFI part from 0x10000 to 0x1FE00
0x10000-10001: Signature = 0x55 0xAA
0x10002: Length = 0x7F -> 0x7F * 0x200 = 0x0FE00
We can now replace the unused space in the legacy vbios with the EFI driver.
Additionally a flag needs to be changed, symbolizing the legacy part is not alone. To find its offset, search for the string "PCIR" and add 17 bytes. I my case it was 0x0023D.
Change it from 0x80 to 0x00.
Last but not least the 8-bit checksum in the last byte of the legacy part needs to be recalculated, which can be done - for example - with the "fixrom" script.
You can find the script and the background for all this information here: http://forum.netkas....59.html#msg3859
I suggest to verify all the changes and the file size by hand to be sure everything is alright.
If you do not know what you are doing, it is better to wait for your vendor to release a hybrid vbios.
I used my XFX legacy vbios and added the EFI part from a similar MSI card.
The XFX legacy vbios got following structure:
1) Legacy part from 0x00000 to 0x0FFFF
0x00000-0x00001: Signature = 0x55 0xAA
0x00002: Length = 0x80 -> 0x80 * 0x200 (block length) = 0x10000
2) Unused space from 0x10000 to 0x1FFFF
The MSI hybrid vbios - on the other hand - looks like this:
1) Legacy part from 0x00000 to 0x0FFFF
0x00000-0x00001: Signature = 0x55 0xAA
0x00002: Length = 0x80 -> 0x80 * 0x200 = 0x10000
2) EFI part from 0x10000 to 0x1FE00
0x10000-10001: Signature = 0x55 0xAA
0x10002: Length = 0x7F -> 0x7F * 0x200 = 0x0FE00
We can now replace the unused space in the legacy vbios with the EFI driver.
Additionally a flag needs to be changed, symbolizing the legacy part is not alone. To find its offset, search for the string "PCIR" and add 17 bytes. I my case it was 0x0023D.
Change it from 0x80 to 0x00.
Last but not least the 8-bit checksum in the last byte of the legacy part needs to be recalculated, which can be done - for example - with the "fixrom" script.
You can find the script and the background for all this information here: http://forum.netkas....59.html#msg3859
I suggest to verify all the changes and the file size by hand to be sure everything is alright.
If you do not know what you are doing, it is better to wait for your vendor to release a hybrid vbios.
Edited by k3nny, 09 May 2013 - 11:33 PM.
#86
Posted 28 February 2013 - 04:47 PM
Ok, I am now pretty damn sure that the problem is a bad implementation of non-UEFI bioses on the 7000-series. I've contacted MSI to try to get a UEFI bios for my card to make sure.
#87
Posted 05 March 2013 - 10:47 PM
Vendor ID 1002Device ID 6819
Subvendor ID 1787
SubDevice ID 201C
Sempre scheda grafica (ma relativa all'audio)
Vendor ID 1002
Device ID AAB0
Subvendor ID 1787
SubDevice ID AAB0
hi in dsdt ?
}Device (GFX0)
{
Name (_ADR, Zero)
Method (_DSM, 4, NotSerialized)
{
Store (Package (0x10)
{
"AAPL,slot-name",
Buffer (0x07)
{
"Slot-1"
},
"device-id",
Buffer (0x04)
{
0x19, 0x68, 0x00, 0x00
},
"@0,name",
Buffer (0x08)
{
"ATY,Aji"
},
"@1,name",
Buffer (0x08)
{
"ATY,Aji"
},
"@2,name",
Buffer (0x08)
{
"ATY,Aji"
},
"@3,name",
Buffer (0x08)
{
"ATY,Aji"
},
"model",
Buffer (0x13)
{
"AMD Radeon HD 7850"
},
"hda-gfx",
Buffer (0x0A)
{
"onboard-1"
}
}, Local0)
DTGP (Arg0, Arg1, Arg2, Arg3, RefOf (Local0))
Return (Local0)
}
}
is correct? thanks
Edited by grillaccio, 06 March 2013 - 06:22 PM.
#88
Posted 06 March 2013 - 07:32 AM
#89
Posted 14 March 2013 - 12:54 PM
When I go to 0x23D on my Sapphire HD7970 the byte is already 0x00 but I see no EFI VBIOS(10000-1FFFF is filled with 0xFFs). I have attached it here if anyone can take a look and see why is that?
Tahiti.rar 39.64K
3 downloads
Now we need a 7970 Mac
Tahiti.rar 39.64K
3 downloads
sea_man, on 06 March 2013 - 07:32 AM, said:
Now we need a 7970 Mac
#90
Posted 14 March 2013 - 01:55 PM
@DAXGr:
Here is the VBIOS I copied the EFI part from for my video card:
msi_7870_hybrid.zip 102.9K
10 downloads
If you compare the structure of both, the offset for the flag is probably 0x241 and is set to 0x80.
Here is the VBIOS I copied the EFI part from for my video card:
msi_7870_hybrid.zip 102.9K
10 downloadsIf you compare the structure of both, the offset for the flag is probably 0x241 and is set to 0x80.
#91
Posted 14 March 2013 - 02:26 PM
k3nny, on 14 March 2013 - 01:55 PM, said:
@DAXGr:
Here is the ROM I copied the EFI part from for my video card:
msi_7870_hybrid.zip
If you compare the structure of both ROMs, the offset for the flag in your ROM is probably 0x241 and is set to 0x80.
Here is the ROM I copied the EFI part from for my video card:
If you compare the structure of both ROMs, the offset for the flag in your ROM is probably 0x241 and is set to 0x80.
EDIT: Wait I see 7870, I have a 7970 would it be a problem?
#92
Posted 14 March 2013 - 02:31 PM
From which VBIOS do you plan to copy the EFI part? That might be interesting for others as well.
EDIT: Most probably it will be. You need to find a 7970 vbios for a card with similar hardware.
EDIT: Most probably it will be. You need to find a 7970 vbios for a card with similar hardware.
#93
Posted 14 March 2013 - 02:36 PM
k3nny, on 14 March 2013 - 02:31 PM, said:
From which VBIOS do you plan to copy the EFI part? That might be interesting for others as well.
EDIT: Most probably it will be. You need to find a 7970 vbios for a card with similar hardware.
EDIT: Most probably it will be. You need to find a 7970 vbios for a card with similar hardware.
#94
Posted 14 March 2013 - 03:49 PM
That one definitely got an EFI driver
Any success yet?
#95
Posted 16 March 2013 - 02:58 PM
grillaccio: can you please share this launch script that you have setup? i got a bit confused at your description, but would like to give it ago if you dont mind sharing!
k3nny: where did you get the efi driver to inject it into the bios? i have an xdx 7970 ghz edition and would like to try patching my bios
k3nny: where did you get the efi driver to inject it into the bios? i have an xdx 7970 ghz edition and would like to try patching my bios
#96
Posted 16 March 2013 - 03:26 PM
@Samawa:
I used an MSI vbios from a card with similar hardware and clocks. The "efi driver" is the "efi part" - just to avoid any misunderstandings.
I used an MSI vbios from a card with similar hardware and clocks. The "efi driver" is the "efi part" - just to avoid any misunderstandings.
#97
Posted 16 March 2013 - 04:23 PM
k3nny: anyway I can convince you to mod mine, for .. you know .. science? Because I really wouldn't know where to start, If it works we can share it on line here and save a lot of people some grief.
http://www.mediafire...4hvwwf4l6zubqdz
http://www.mediafire...4hvwwf4l6zubqdz
#98
Posted 16 March 2013 - 05:08 PM
For science it is then:
xfx_hd7970_ghz.zip 473.5K
12 downloads
The efi part is from an MSI R7970-2PMD3GD5/OC.
Flashing at your own risk in DOS with "atiflash -f -p 0 new.rom".
After a reboot you can see if atiflash recognizes the new vbios with "atiflash.exe -ai 0"
xfx_hd7970_ghz.zip 473.5K
12 downloadsThe efi part is from an MSI R7970-2PMD3GD5/OC.
Flashing at your own risk in DOS with "atiflash -f -p 0 new.rom".
After a reboot you can see if atiflash recognizes the new vbios with "atiflash.exe -ai 0"
#99
Posted 16 March 2013 - 05:20 PM
Will do this now! thanks k3nny will report back, for science!
#100
Posted 16 March 2013 - 05:41 PM
k3nny, on 16 March 2013 - 05:08 PM, said:
For science it is then:
xfx_hd7970_ghz.zip
The efi part is from an MSI R7970-2PMD3GD5/OC.
Flashing at your own risk in DOS with "atiflash -f -p 0 new.rom".
After a reboot you can see if atiflash recognizes the new vbios with "atiflash.exe -ai 0"
The efi part is from an MSI R7970-2PMD3GD5/OC.
Flashing at your own risk in DOS with "atiflash -f -p 0 new.rom".
After a reboot you can see if atiflash recognizes the new vbios with "atiflash.exe -ai 0"
K3nny does this patch resolve the sleep trick?
0 user(s) are reading this topic
0 members, 0 guests, 0 anonymous users



Sign In
Create Account








