ProjectOSX

Welcome Guest!

Returning User? Login here.

Want membership privileges? Register here.

707 Pages V  « < 289 290 291 292 293 > »   
Start a new topic Add Reply
> Clover
Krazubu
post Mar 8 2013, 06:59 PM
Post #5801
Hey guys,

I've been following this project for a while and I'm glad to see it gets mature.
I just post my 2 cents for feedback.
As many people, I also get sleep issues (only with Clover, always works fine with XPC or chameleon for instance).
The problem is the same using either the MBR or UEFI one.
From times to times it won't wake up. It doesn't reboot, it's just "locked in black" but it seems badly crashed.
It works fine other times.
eurisko
post Mar 8 2013, 07:36 PM
Post #5802
QUOTE (Krazubu @ Mar 8 2013, 07:59 PM) *
Hey guys,

I've been following this project for a while and I'm glad to see it gets mature.
I just post my 2 cents for feedback.
As many people, I also get sleep issues (only with Clover, always works fine with XPC or chameleon for instance).
The problem is the same using either the MBR or UEFI one.
From times to times it won't wake up. It doesn't reboot, it's just "locked in black" but it seems badly crashed.
It works fine other times.



Welcome to the club....See this post, it may be useful for you: http://www.projectosx.com/forum/index.php?showtopic=2565

This post has been edited by eurisko: Mar 8 2013, 07:37 PM
GA-Z68X-UD4-B3 (Rev. 1.0) UEFI BIOS U1e- INTEL I5-2500K- ASUS EN9800GT 1GB VRAM- 16 GB RAM DDR3 1600 CORSAIR VENGEANCE BLUE EDITION
SSD OCZ VERTEX 4 120 GB (MLION)- SSD OCZ VERTEX 3 120 GB (WINDOWS 8 X64)- HDD WD CAVIAR BLACK 1TB (DATA)- MLION 10.8.4 (12E55)
WINDOWS 8 PRO x64
magnifico
post Mar 9 2013, 03:24 PM
Post #5803
Hi dev ...and all friend ...
I have a constant problem when you first start the computer, only the first boot, gets stuck on dsmos arrived, then I have to shut down forcibly and at the next startup everything 'ok....help please
Asrock z87 oc formula / i7 4770k @4,7 Ghz \SSD Samsung 840 Pro 128 GB\ 2x128GB SuperSSpeed S301(SLC) RAID OSX Software\ Samsung 840 EVO 120 GB\Nvidia 660 Gigabyte OC \Corsair Vengeance 4x4GB (16GB) DDR3 1600
CloverUEFI mode [/b] Score GeekBench said the flea nut to give me time and I will pound
STLVNUB
post Mar 9 2013, 05:32 PM
Post #5804
CloverGrower can now build rEFInd, in case anyone interested.
use refined.sh

This post has been edited by STLVNUB: Mar 9 2013, 05:33 PM
Azrock Z77 Pro3, G1610 AMD HD7750 Clover/Ozmosis
Toshiba A660-0MR00R I7-7740QM GT330M 8 Gig Clover

STLVNUB
post Mar 9 2013, 09:19 PM
Post #5805
revision 1159 is broken, exit stage left, and let the games begin...

This post has been edited by STLVNUB: Mar 9 2013, 09:29 PM
Azrock Z77 Pro3, G1610 AMD HD7750 Clover/Ozmosis
Toshiba A660-0MR00R I7-7740QM GT330M 8 Gig Clover

apianti
post Mar 9 2013, 09:40 PM
Post #5806
Truthfully, those last two commits are disasters. Breaking the build script is not good. I also hate how it makes you use a different styling of arguments than the EDK build system. It's stupidly confusing and makes no sense. Why is --gcc47 easier than -t GCC47? --ia32 than -a IA32? Basically all of the options for the script are the same but makes it so you can't change additional options available to you. It's dumb.

EDIT: Some like --clean (compared to clean) are actually longer...

This post has been edited by apianti: Mar 9 2013, 09:43 PM
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3
STLVNUB
post Mar 9 2013, 10:01 PM
Post #5807
QUOTE (apianti @ Mar 10 2013, 07:40 AM) *
Truthfully, those last two commits are disasters. Breaking the build script is not good. I also hate how it makes you use a different styling of arguments than the EDK build system. It's stupidly confusing and makes no sense. Why is --gcc47 easier than -t GCC47? --ia32 than -a IA32? Basically all of the options for the script are the same but makes it so you can't change additional options available to you. It's dumb.

EDIT: Some like --clean (compared to clean) are actually longer...

+1
My CloverGrower just died a slow and painful death, so much for the 'Pro' blink.gif

edit:
Talking about 'Pro', just trying that now.
Bet you it DON'T work.


This post has been edited by STLVNUB: Mar 9 2013, 10:22 PM
Azrock Z77 Pro3, G1610 AMD HD7750 Clover/Ozmosis
Toshiba A660-0MR00R I7-7740QM GT330M 8 Gig Clover

JrCs
post Mar 9 2013, 10:25 PM
Post #5808
QUOTE (apianti @ Mar 9 2013, 10:40 PM) *
Truthfully, those last two commits are disasters. Breaking the build script is not good. I also hate how it makes you use a different styling of arguments than the EDK build system. It's stupidly confusing and makes no sense. Why is --gcc47 easier than -t GCC47? --ia32 than -a IA32? Basically all of the options for the script are the same but makes it so you can't change additional options available to you. It's dumb.

EDIT: Some like --clean (compared to clean) are actually longer...

All the old options work as before.
EDK build system haven't a specific styling pf arguments. It's use the build script that have (like any other command on unix) short (with one -) and long (with two --) arguments.
We can change the options to be more like unix commands (that i'll prefer), but in my commit i only convert long options of ebuild.sh: -ia32 to --ia32. But like i said old options always works for compatibility.

The target of the commit wasn't to change options but clean the ebuild.sh script.

This post has been edited by JrCs: Mar 9 2013, 10:33 PM
JrCs
post Mar 9 2013, 10:28 PM
Post #5809
QUOTE (STLVNUB @ Mar 9 2013, 11:01 PM) *
+1
My CloverGrower just died a slow and painful death, so much for the 'Pro' blink.gif

edit:
Talking about 'Pro', just trying that now.
Bet you it DON'T work.

Don't understand STLVNUB.
I keep the old options for compatibility. So the interface didn't change between CloverGrower[Pro] and ebuild.sh
apianti
post Mar 9 2013, 10:32 PM
Post #5810
@JrCs,

I know I can be a huge dick sometimes, and not an attack on you or anything but you have to fix the build script, like today. Another thing is that the custom rc script can't change names, it has to be hardcoded just like /etc/rc.local is executed always from /etc/rc.local. If you change the script name randomly between revisions it defeats the purpose of the custom script in the first place which is to prevent unnecessary installation procedures every time you update. Also why does it need another log? Appending it to the current log is perfectly acceptable and probably preferable to clutter. In fact it seems it will append it to both the current log and a new custom log. Also seems you changed log every boot, to also include the system log, which pretty much is just extra non sense because you have the full copy of every boot. Setting the line count to zero used to use the whole previous log, now it does nothing in that case and still trims the current log lines instead of the previous.
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3
apianti
post Mar 9 2013, 10:35 PM
Post #5811
QUOTE (JrCs @ Mar 9 2013, 05:25 PM) *
All the old options work as before.
EDK build system haven't a specific styling pf arguments. It's use the build script that have (like any other command on unix) short (with one -) and long (with two --) arguments.
We can change the options to be like unix commands (that i'll prefer), but in my commit i only convert (but keep) long options of ebuild.sh: -ia32 to --ia32

The target of the commit wasn't to change options but clean the ebuild.sh script.

EDK definitely has a specific styling of arguments, go to you edk folder, run edksetup and type build --help. Look at the windows build script it uses the exact edk style, so that you can pass in any argument edk build accepts. You can't do that through ebuild. Also I know you didn't change the options I was just pointing out that they were slightly different for no good reason and actually cripple the ability to use the edk build system.
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3
JrCs
post Mar 9 2013, 10:40 PM
Post #5812
QUOTE (apianti @ Mar 9 2013, 11:32 PM) *
@JrCs,

I know I can be a huge dick sometimes, and not an attack on you or anything but you have to fix the build script, like today. Another thing is that the custom rc script can't change names, it has to be hardcoded just like /etc/rc.local is executed always from /etc/rc.local. If you change the script name randomly between revisions it defeats the purpose of the custom script in the first place which is to prevent unnecessary installation procedures every time you update.

I don't change the name of the script ?


QUOTE (apianti @ Mar 9 2013, 11:32 PM) *
Also why does it need another log? Appending it to the current log is perfectly acceptable and probably preferable to clutter. In fact it seems it will append it to both the current log and a new custom log.

Because the script are for 2 differents purpose (like rc.local and rc.shutdown.local). And it's easier to debug than putting all in one file

QUOTE (apianti @ Mar 9 2013, 11:32 PM) *
Also seems you changed log every boot, to also include the system log, which pretty much is just extra non sense because you have the full copy of every boot. Setting the line count to zero used to use the whole previous log, now it does nothing in that case and still trims the current log lines instead of the previous.

I don't change that things in the last commits i do today
STLVNUB
post Mar 9 2013, 10:41 PM
Post #5813
QUOTE (JrCs @ Mar 10 2013, 08:28 AM) *
Don't understand STLVNUB.
I keep the old options for compatibility. So the interface didn't change between CloverGrower[Pro] and ebuild.sh

TOOLCHAIN is exported in CloverGrower[Pro]; this is all that is needed.

In ebuild.sh It is also exported as GCC47 and defined with -gcc47 when compiling it looks for /GCC47/X86 and FAILS.

edit:
looks like bug is in CloverGrower.


This post has been edited by STLVNUB: Mar 9 2013, 10:47 PM
Azrock Z77 Pro3, G1610 AMD HD7750 Clover/Ozmosis
Toshiba A660-0MR00R I7-7740QM GT330M 8 Gig Clover

apianti
post Mar 9 2013, 10:45 PM
Post #5814
QUOTE (JrCs @ Mar 9 2013, 05:40 PM) *
I don't change the name of the script ?

You made the ability to change the name of the script, when I had hard coded it in. It was an unnecessary change.
QUOTE (JrCs @ Mar 9 2013, 05:40 PM) *
Because the script are for 2 differents purpose (like rc.local and rc.shutdown.local). And it's easier to debug than putting all in one file

This makes no sense. Why do I need it logged into two files? Actually four once all scripts are run. custom.rc.local logs to rc.local.log already and doesn't need it's own log. Same for the other.
QUOTE (JrCs @ Mar 9 2013, 05:40 PM) *
I don't change that things in the last commits i do today

Pointed those changes out when you originally modified my scripts. You've never fixed them.
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3
JrCs
post Mar 9 2013, 10:46 PM
Post #5815
QUOTE (apianti @ Mar 9 2013, 11:35 PM) *
EDK definitely has a specific styling of arguments, go to you edk folder, run edksetup and type build --help. Look at the windows build script it uses the exact edk style, so that you can pass in any argument edk build accepts. You can't do that through ebuild. Also I know you didn't change the options I was just pointing out that they were slightly different for no good reason and actually cripple the ability to use the edk build system.

STOP !
I was cleaning ebuild.sh, i don't try to make the arguments of ebuild.sh the same as EDK build script.
We can do that in the futur, but my priority wasn't to mimic EDK II build script.
My priority is to have the possibility to add build options specific to Clover, like activating VBIOS Patch in CloverEFI, use BiosBlockIO instead of sata, etc...
JrCs
post Mar 9 2013, 10:50 PM
Post #5816
QUOTE (apianti @ Mar 9 2013, 11:45 PM) *
You made the ability to change the name of the script, when I had hard coded it in. It was an unnecessary change.

By creating a variable ?
It's not to made the ability to change the name, but to make the script more easier to maintain than copy/paste hard coded path (which is the best way to generate bugs) !

QUOTE (apianti @ Mar 9 2013, 11:45 PM) *
This makes no sense. Why do I need it logged into two files? Actually four once all scripts are run. custom.rc.local logs to rc.local.log already and doesn't need it's own log. Same for the other.

If you prefer we can remove the creation of a new log

QUOTE (apianti @ Mar 9 2013, 11:45 PM) *
Pointed those changes out when you originally modified my scripts. You've never fixed them.

So i will look at it.

This post has been edited by JrCs: Mar 9 2013, 10:52 PM
apianti
post Mar 9 2013, 10:52 PM
Post #5817
QUOTE (JrCs @ Mar 9 2013, 05:46 PM) *
STOP !
I was cleaning ebuild.sh, i don't try to make the arguments of ebuild.sh the same as EDK build script.
We can do that in the futur, but my priority wasn't to mimic EDK II build script.
My priority is to have the possibility to add build options specific to Clover, like activating VBIOS Patch in CloverEFI, use BiosBlockIO instead of sata, etc...

I know this, I wasn't even talking to you or about you during that part, and just said that in the comment to which you're replying. And OMG, NO, DO NOT DO THAT, THAT IS A HORRIFIC IDEA. That's what DSC, DEC, and INF files are for, and if the script mimicked edk's build options wouldn't even need to be done because you could just specify building another DSC, FDF, or INF.
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3
apianti
post Mar 9 2013, 11:00 PM
Post #5818
QUOTE (JrCs @ Mar 9 2013, 05:50 PM) *
By creating a variable ?
It's not to made the ability to change the name, but to make the script more easier to maintain than copy/paste hard coded path (which is the best way to generate bugs) !

laugh.gif Actually the most common cause of bugs is incorrect use of variables. Look it up.
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3
STLVNUB
post Mar 9 2013, 11:19 PM
Post #5819
Fixed bugs in CloverGrower ( back to using sed for tools_def.txt), now lets see how it handles latest commits.

edit:
more bug fixes, now working

This post has been edited by STLVNUB: Mar 10 2013, 12:30 AM
Azrock Z77 Pro3, G1610 AMD HD7750 Clover/Ozmosis
Toshiba A660-0MR00R I7-7740QM GT330M 8 Gig Clover

apianti
post Mar 9 2013, 11:36 PM
Post #5820
@JrCs,

I'm pretty sure you broke all the build scripts now..... Windows is for sure broken.
Gigabyte GA-Z68XP-UD3P Firmware U1g, Intel Core i5-2500K, Corsair Vengeance 4x4GB (16GB) DDR3 1600, XFX Radeon 6870 2GB,
Western Digital Caviar Black 1TB, LG Rewritable BluRay Disc Burner, D-Link DWA-556 Wireless N Xtreme PCIe

Windows 8.1 Pro, Mac OS X 10.9 Mavericks, Ubuntu 13.10, LinuxMINT 15 Cinnamon, Fedora 19, openSUSE 12.3

707 Pages V  « < 289 290 291 292 293 > » 
Add Reply Start a new topic
7 User(s) are reading this topic (7 Guests and 0 Anonymous Users)
0 Members: