Clover General discussion

[Slice]

9 hours ago, Jief_Machak said:

@Matgen84 Could you test (BiosVersion, ReleaseData) with this version ?

@Slice No idea why it would stop between  "CurrentMode: Width=1920 Height=1080" and "SetNvramVariable (FirmwareFeaturesMask, guid, 0x6, 4):" I've added few logs. Could you re-test when you can ?

Thanks.
 CloverX64-2021-04-22-12-45-56-a1a27d2-dirty-jief.zip

Same stop

 

2021-4-22_19-11_CLOVERX64.EFI.log.zip

[PG7]

hello team

 

I am always following the good path of Clover and I always testify when time allows me

 

I tested it last night and today the boot that Jief disposes here in an

All In One Asus A4310-BB020T
intel core i3-4160T
intel HD4400 GT1,5 (just the 'apercu' of the photos I can't use because of this GT1,5 but here I have an i5-4590T to put and move to the HD4600 GT2)
8 go RAM
HDD 500 go 7200rpm (maybe change to SSD)

the boot and usage tests work perfect here

I leave my thanks to you for always continuing this fight and putting my favorite bootloader better and better with you only this is possible

I will always be here for the tests even if I don’t answer it doesn’t mean I’ve abandoned the opposite always present

if at a moment there is a lack of tests on a particular machine that I have, just tell me that I will do it

 

good continuation and strength that CLOVER is always in high shape

 

Google translate !

 

Spoiler

 

[Jief_Machak]

Added even more log. This time, it seems impossible to not have at least one line more !

CloverX64-2021-04-23-00-35-03-a1a27d2-dirty-jief.zip

[Slice]

5 hours ago, Jief_Machak said:

Added even more log. This time, it seems impossible to not have at least one line more !

CloverX64-2021-04-23-00-35-03-a1a27d2-dirty-jief.zip 826.29 kB · 9 downloads

One more line

 

2021-4-23_3-0_CLOVERX64.EFI.log.zip

[Jief_Machak]

3 hours ago, Slice said:

One more line

 

2021-4-23_3-0_CLOVERX64.EFI.log.zip 9.51 kB · 2 downloads

Some more logs !CloverX64-2021-04-23-10-10-23-a1a27d2-dirty-jief.zip

[naiclub]

Hello everyone Can anyone tell which version of Clover has been working?

Edited April 23, 2021 by naiclub

[Jief_Machak]

Define "working" 

[naiclub]

What happened? Now built at 5129.

[Jief_Machak]

44 minutes ago, naiclub said:

What happened? Now built at 5129.

Fixed.

17 minutes ago, MifJpn said:

as information.
In fact, if you edit config.plist with
<true/> is replaced with <true />
<false/> will be replaced with <false />.
In fact, I don't know how much is acceptable as correct XML.
However, OpenCore seems to allow it, but in the case of Clover, the following display appears and it does not work.

To me, PlistEDplus seems to use a dialect.
Therefore, I think that PlistEDplus will not be able to support a lot of XML, so I will tell the author.
I will write it as information.

Sorry.

It is fix at new release1.0.50!!
Thank you.

That's interesting.

Panicking because of a problem in config.plist is a sign a technical bug somewhere. Wrong input must ignored, corrected or whatever, but not cause a panic.

Could you send me this config.plist that causes this bug ?

[Matgen84]

@Jief_Machak

I've got problem with cloverbootloader/opencorepkg commit version:
 

iMac-de-Mathieu:cloverbootloader mathieu$ git pull --recurse-submodules origin master
From https://github.com/CloverHackyColor/CloverBootloader
 * branch                master     -> FETCH_HEAD
Fetching submodule OpenCorePkg
Already up to date.
Submodule path 'OpenCorePkg': checked out '444a41b15eee65123be0b47b2f4813c6b11a0520'

In the repo, CloverHackycolor/Cloverbootloader/Opencorepkg ---->   commit   444a41b15eee65123be0b47b2f4813c6b11a0520

In the repo, CloverHackycolor/Opencorepkg ---->  commit    2378f2b5ee97ed31333d49881620af447e2b3095

Do you have a solution !

[Jief_Machak]

16 minutes ago, MifJpn said:

Thank you for picking it up.
This is it.

config.plist 39.01 kB · 2 downloads

If you actually replace ”true /” with ”true/” and ”false /” with ”false/” in CotEditer, it will work.

Thank you.

Sorry, my bad. The problem is with the plist of the theme. Could you send me a zip of the theme, preferably with the problem.

9 minutes ago, Matgen84 said:

@Jief_Machak

I've got problem with cloverbootloader/opencorepkg commit version:
 

iMac-de-Mathieu:cloverbootloader mathieu$ git pull --recurse-submodules origin master
From https://github.com/CloverHackyColor/CloverBootloader
 * branch                master     -> FETCH_HEAD
Fetching submodule OpenCorePkg
Already up to date.
Submodule path 'OpenCorePkg': checked out '444a41b15eee65123be0b47b2f4813c6b11a0520'

In the repo, CloverHackycolor/Cloverbootloader/Opencorepkg ---->   commit   444a41b15eee65123be0b47b2f4813c6b11a0520

In the repo, CloverHackycolor/Opencorepkg ---->  commit    2378f2b5ee97ed31333d49881620af447e2b3095

Do you have a solution !

cd OpencorePkg

git checkout master

git fetch

git pull

cd ..

 

[Matgen84]

1 minute ago, Jief_Machak said:

Sorry, my bad. The problem is with the plist of the theme. Could you send me a zip of the theme, preferably with the problem.

cd OpencorePkg

git checkout master

git fetch

git pull

cd ..

 

 

Thanks 

What do you think about:
 

In the GitHub repo, CloverHackycolor/Cloverbootloader/Opencorepkg ---->   commit   444a41b15eee65123be0b47b2f4813c6b11a0520

In the GitHub repo, CloverHackycolor/Opencorepkg ---->  commit    2378f2b5ee97ed31333d49881620af447e2b3095

[Jief_Machak]

4 minutes ago, Matgen84 said:

 

Thanks 

What do you think about:
 

 

In the GitHub repo, CloverHackycolor/Cloverbootloader/Opencorepkg ---->   commit   444a41b15eee65123be0b47b2f4813c6b11a0520

In the GitHub repo, CloverHackycolor/Opencorepkg ---->  commit    2378f2b5ee97ed31333d49881620af447e2b3095

Yes, it was forgotten to commit in Clover the fact that there is a new commit in OpenCorePkg.

done.

Thanks for reporting.

[Matgen84]

51 minutes ago, Jief_Machak said:

Yes, it was forgotten to commit in Clover the fact that there is a new commit in OpenCorePkg.

done.

Thanks for reporting.

Thanks.
Few days ago, I've a discussion about this.
Is it possible to automate to commit in Clover the fact that there is a new commit in OpenCorePkg !

EDIT 1: now this (sorry, maybe, I am a little bit stupid)

 

Entering 'OpenCorePkg'
HEAD is now at 2378f2b5 CpuScan delegated to Clover
HEAD is now at afa09ccb4 OpenCorePkg new commit.

iMac-de-Mathieu:opencorepkg mathieu$ git checkout master
Previous HEAD position was 2378f2b5 CpuScan delegated to Clover
Switched to branch 'master'
Your branch is behind 'origin/master' by 2 commits, and can be fast-forwarded.
  (use "git pull" to update your local branch)
iMac-de-Mathieu:opencorepkg mathieu$ git fetch
iMac-de-Mathieu:opencorepkg mathieu$ git pull
Updating 77af4483..2378f2b5
Fast-forward
 Include/Intel/IndustryStandard/ProcessorInfo.h | 4 ++--
 Library/OcCpuLib/AppleCpuSupport.c             | 8 ++++----
 Platform/OpenCore/OpenCore.c                   | 4 ++--
 3 files changed, 8 insertions(+), 8 deletions(-)

EDIT 2: Normally, I use BuildMe script, menu option 13

 

updateResetClover() {
echo "[UPDATE RESET CLOVER]"
cd "${CLOVERROOT}"
if [[ -d .git ]]; then
  git fetch --all --recurse-submodules
  git submodule foreach --recursive git clean -d -f -f
  git clean -d -f -f
  git submodule foreach --recursive git reset --hard origin/master
  git reset --hard origin/master
  git pull --recurse-submodules origin master
else
  echo "Error: this is not a git repository, can't update!"
fi
}

 

Edited April 23, 2021 by Matgen84

[Slice]

8 hours ago, Jief_Machak said:

Some more logs !CloverX64-2021-04-23-10-10-23-a1a27d2-dirty-jief.zip

Some more lines

 

2021-4-23_15-35_CLOVERX64.EFI.log.zip

[Slice]

Anyway my own compilation of latest commit works. 

afa09ccb4eb349f9a0d1cfbfbfde2d9b7a57b4eb

[Jief_Machak]

1 hour ago, Slice said:

Some more lines

 

2021-4-23_15-35_CLOVERX64.EFI.log.zip 9.59 kB · 1 download

Are we facing a compiler problem ?

Because from your log, StrToGuidLE is called with a null value, from that line 

EFI_STATUS Status = StrToGuidLE(Smbios.SmUUID, uuid); 

in settings.cpp from method 

const XString8& SETTINGS_DATA::getUUID(EFI_GUID *uuid)

but Smbios.SmUUID cannot be null : it's not a pointer.

 

Could you try this : CloverX64-2021-04-23-20-04-58-afa09cc-dirty-jief-gcc.zip it's a gcc compilation, and, if it panic, this : CloverX64-2021-04-23-20-04-58-afa09cc-dirty-jief-xcode.zip   ?

 

Thanks.

 

[Slice]

2 hours ago, Jief_Machak said:

Are we facing a compiler problem ?

Because from your log, StrToGuidLE is called with a null value, from that line 

EFI_STATUS Status = StrToGuidLE(Smbios.SmUUID, uuid); 

in settings.cpp from method 

const XString8& SETTINGS_DATA::getUUID(EFI_GUID *uuid)

but Smbios.SmUUID cannot be null : it's not a pointer.

 

Could you try this : CloverX64-2021-04-23-20-04-58-afa09cc-dirty-jief-gcc.zip it's a gcc compilation, and, if it panic, this : CloverX64-2021-04-23-20-04-58-afa09cc-dirty-jief-xcode.zip   ?

 

Thanks.

 

Both are not working. CPU stops or infinite loop.

But there is something new before start

[Jief_Machak]

58 minutes ago, MifJpn said:

Follwing is what I understand.
Reference:　https://www.w3.org/TR/xml/
<true></true> = <true/>
The end of "Tags for empty element" is "/>".
Therefore, the space before ”/>” is probably acceptable.

 

It seems to be arbitrary in XML whether to put a space before />, but in W3C, as follows
https://www.w3.org/TR/xhtml1/#guidelines
It seems that it is supposed to put a space.

Thank you.

Could you send me the theme you were using when that panic happened ?

1 hour ago, Slice said:

Both are not working. CPU stops or infinite loop.

But there is something new before start

Both not working : should not be a compiler issue.

Could you send a debug log, if you have one ?

What I don't understand is why I can't reproduce that here with your config.plist...

 

Yes, for now I display all errors and warnings from config.plist. That'll push people to correct them ! 

 

Could you reproduce that in Qemu ?

[Slice]

8 hours ago, Jief_Machak said:

Could you send me the theme you were using when that panic happened ?

Both not working : should not be a compiler issue.

Could you send a debug log, if you have one ?

What I don't understand is why I can't reproduce that here with your config.plist...

 

Yes, for now I display all errors and warnings from config.plist. That'll push people to correct them ! 

 

Could you reproduce that in Qemu ?

Yes, reproduced in Qemu. First one (gcc compilation)

custom properties for device 00:01.00 injected
custom properties for device 00:03.00 injected
custom properties for device 00:1D.00 injected
 RCBA disabled; cannot use it
stringlength = 1620
CurrentMode: Width=1024 Height=768
SetVariablesForOSX -> Enter
SetVariablesForOSX -> getUUID
getUUID -> 0
getUUID -> 3
StrToGuidLE -> 0
StrToGuidLE -> 1  0
getUUID -> 4
SmUUID() is not valid
A fatal error happened. System halted.

This is my old qemu with old img.

config.plist.zip

I also can't reproduce how can you create such binaries 

[Matgen84]

15 hours ago, Matgen84 said:

Thanks.
Few days ago, I've a discussion about this.
Is it possible to automate to commit in Clover the fact that there is a new commit in OpenCorePkg !

EDIT 1: now this (sorry, maybe, I am a little bit stupid)

 

Entering 'OpenCorePkg'
HEAD is now at 2378f2b5 CpuScan delegated to Clover
HEAD is now at afa09ccb4 OpenCorePkg new commit.

iMac-de-Mathieu:opencorepkg mathieu$ git checkout master
Previous HEAD position was 2378f2b5 CpuScan delegated to Clover
Switched to branch 'master'
Your branch is behind 'origin/master' by 2 commits, and can be fast-forwarded.
  (use "git pull" to update your local branch)
iMac-de-Mathieu:opencorepkg mathieu$ git fetch
iMac-de-Mathieu:opencorepkg mathieu$ git pull
Updating 77af4483..2378f2b5
Fast-forward
 Include/Intel/IndustryStandard/ProcessorInfo.h | 4 ++--
 Library/OcCpuLib/AppleCpuSupport.c             | 8 ++++----
 Platform/OpenCore/OpenCore.c                   | 4 ++--
 3 files changed, 8 insertions(+), 8 deletions(-)

EDIT 2: Normally, I use BuildMe script, menu option 13

 

updateResetClover() {
echo "[UPDATE RESET CLOVER]"
cd "${CLOVERROOT}"
if [[ -d .git ]]; then
  git fetch --all --recurse-submodules
  git submodule foreach --recursive git clean -d -f -f
  git clean -d -f -f
  git submodule foreach --recursive git reset --hard origin/master
  git reset --hard origin/master
  git pull --recurse-submodules origin master
else
  echo "Error: this is not a git repository, can't update!"
fi
}

 

Hi @Jief_Machak I know that you're busy at this moment, but if you've few minutes to take look to my comment posted 15 hours ago. Please. It would be great
 

[Jief_Machak]

3 minutes ago, Matgen84 said:

Is it possible to automate to commit in Clover

Not sure we want to automate commits. But sure, we could put a warning in ebuild.sh, for example...

 

4 minutes ago, Matgen84 said:

git fetch --all --recurse-submodules git submodule foreach --recursive git clean -d -f -f git clean -d -f -f git submodule foreach --recursive git reset --hard origin/master git reset --hard origin/master git pull --recurse-submodules origin master

We probably should add a git checkout master somewhere here to always pull the right OpencorePkg version that goes with the current Clover commit.

It's not the same thing as automating commits...

And to test that, I've either to setup test repositories, or wait for the next OpencorePkg commit. I will do something, but not top of the list at the moment.

[Jief_Machak]

@Slice Ok, got the bug, finally. Bad initialisation of smUUID in the new parser. Strange thing is that you didn't get the panic message at first.

@MifJpn Could you recreate the situation where you got the panic and send me a zip of your whole CLOVER folder ?

[Matgen84]

1 hour ago, Jief_Machak said:

Not sure we want to automate commits. But sure, we could put a warning in ebuild.sh, for example...

 

We probably should add a git checkout master somewhere here to always pull the right OpencorePkg version that goes with the current Clover commit.

It's not the same thing as automating commits...

And to test that, I've either to setup test repositories, or wait for the next OpencorePkg commit. I will do something, but not top of the list at the moment.

 

Thanks @Jief_Machak. That would be good indeed: "add a git checkout master somewhere here to always pull the right OpencorePkg version that goes with the current Clover commit." You're right, also: "not top of the list at the moment".

[Matgen84]

13 hours ago, MifJpn said:

Follwing is what I understand.
Reference:　https://www.w3.org/TR/xml/
<true></true> = <true/>
The end of "Tags for empty element" is "/>".
Therefore, the space before ”/>” is probably acceptable.

 

It seems to be arbitrary in XML whether to put a space before />, but in W3C, as follows
https://www.w3.org/TR/xhtml1/#guidelines
It seems that it is supposed to put a space.

Thank you.

Hi Bro @MifJpn Just my opinion, there is a bug with PlistEdPlus since 1.0.50: the version introduce a space before "/>". I've got no issues with BBedit, PlisEdit Pro or SublimeText (no space).
The W3C guideLines for HTML, if I understand well: 'For empty element, also, use the minimized tag syntax for empty elements, e.g. <br />, as the alternative syntax <br></br> allowed by XML gives uncertain results in many existing user agents.' It's only for web HTML pages.

Clover sample.plist or Opencore config.plist don't use the space before ”/>”