It is currently Tue Sep 02, 2014 8:40 pm

All times are UTC - 8 hours




Post new topic Reply to topic  [ 60 posts ]  Go to page 1, 2, 3  Next
Author Message
 Post subject: Explorer's modification of CPUstress image
PostPosted: Mon Feb 25, 2013 11:15 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
This thread is for discussion of my (Explorer09's) modification of the CPUstress image.

The original, Icecube's image can be found in this thread. After fixing the annoying kernel panic problem, I keep improving the image (including updating programs and fixing more bugs).

The original bug-reporting thread has been filled with off-topic discussions since then, so I start this thread to clean things up.

Latest version: CPUstress 2.5.1 (20 August 2014)
https://dl.dropboxusercontent.com/u/701 ... s-2.5.1.7z
Size: 4626848 bytes
MD5: cd70fdcc52ff906399e3944909042995
SHA1: 962af04a85e43d5646db86687718c1fde784cdb9

Image

Development page on GitHub:
https://github.com/Explorer09/cpustress-ubcd

Programs included in the CPUstress image v2.5.1:


Last edited by Explorer09 on Wed Aug 20, 2014 12:35 am, edited 27 times in total.

Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Mon Feb 25, 2013 11:52 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
Here is the content of the "help files":
Quote:
>> PAGE 1

You can press [TAB] to edit the boot options in isolinux.

For CPU Burn-in, you can specify the time it runs by using either 'days',
'hours', or 'minutes' option. By default, CPU Burn-in runs for 7 days
(=10080 minutes), which is equivalent to this:

> .com32 linux.c32 /pmagic/bzImage root=/dev/ram0 ramdisk_size=12000 noapic
quiet ubcdcmd=cpuburn initrd=/ubcd/boot/cpustress/initrd.gz days=7

If more than one of the time options are specified, the run time will be
determined by the value of larger unit (days > hours > minutes) and the rest
are ignored.

For Mesenne Prime Test, you can run the older v23.9.2 instead of v27.9 by
replacing "mprime27" with "mprime23" in the 'ubcdcmd' option:

> .com32 linux.c32 /pmagic/bzImage root=/dev/ram0 ramdisk_size=12000 noapic
quiet ubcdcmd=mprime23 initrd=/ubcd/boot/cpustress/initrd.gz

<< Press [F2] to continue to next page, any other key to return to menu >>

>> PAGE 2

You can press [TAB] to edit the boot options in isolinux.

For Stress and StressCPU, you may pass the command-line arguments to the
program by using the 'cmdstress' and 'cmdstresscpu2' options respectively in isolinux. Examples:

> .com32 linux.c32 /pmagic/bzImage root=/dev/ram0 ramdisk_size=12000 noapic
quiet ubcdcmd=stress cmdstress="--quiet -c 1k"
initrd=/ubcd/boot/cpustress/initrd.gz

> .com32 linux.c32 /pmagic/bzImage root=/dev/ram0 ramdisk_size=12000 noapic
quiet ubcdcmd=stresscpu2 cmdstresscpu2="-s -t 48:00:00"
initrd=/ubcd/boot/cpustress/initrd.gz

For more information, please run the helpinfo program in the CPUstress image.
You can run helpinfo directly from isolinux by specifying 'ubcdcmd=helpinfo':

> .com32 linux.c32 /pmagic/bzImage root=/dev/ram0 ramdisk_size=12000 noapic
quiet ubcdcmd=helpinfo initrd=/ubcd/boot/cpustress/initrd.gz

<< Press [F1] to return to previous page, any other key to return to menu >>


Generally I just document the "boot cheat codes" that are already in the CPUstress image.
They are in the /help/syslinux/ folder in my cpustress-2.1expl6 archive.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 7:06 am 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
Thank you for sharing you improvements.

Just to show one possible way to handle the options, I am attaching 2 screenshots.

The specific aesthetics of the boot menu is less important.

In the “first step”, the user chooses which test to run (one of the tests, or cpuinfo, or help, or tips about possible commands).

After the user presses [ENTER] on the selected test, a second selection appears, which only corresponds to the options offered by that specific test.

All is done in one cfg file only, optionally adding the two DISPLAY files that Explorer09 already prepared.

As said, there are other ways. This is just one possibility (with pros and cons).

Attachment:
cpustress1.jpg
cpustress1.jpg [ 17.9 KiB | Viewed 19590 times ]

Attachment:
cpustress2.jpg
cpustress2.jpg [ 15.44 KiB | Viewed 19590 times ]


Now, regarding the help, ubcdcmd=helpinfo, cmdhelpinfo=* (and I may be forgetting some other option), it seems a little bit confusing. Reading the help files, it is not so clear why there is a need for both "ubcdcmd=helpinfo" and "cmdhelpinfo=*", and that's in addition to the "ubcdcmd=menu" option.

Once again, thank you Explorer09 8) .
Best Regards,
Ady.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 5:07 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
ady wrote:
After the user presses [ENTER] on the selected test, a second selection appears, which only corresponds to the options offered by that specific test.
All is done in one cfg file only, optionally adding the two DISPLAY files that Explorer09 already prepared.
As said, there are other ways. This is just one possibility (with pros and cons).

Some programs, such as Stress, can have many many options to test.

Your idea can work if there's only a few options (or presets) to use. However, none of the program developers offer me such presets.

ady wrote:
Now, regarding the help, ubcdcmd=helpinfo, cmdhelpinfo=* (and I may be forgetting some other option), it seems a little bit confusing. Reading the help files, it is not so clear why there is a need for both "ubcdcmd=helpinfo" and "cmdhelpinfo=*", and that's in addition to the "ubcdcmd=menu" option.


The "cmdhelpinfo" option is to specify which help to display inside the "helpinfo" script. For example, "ubcdcmd=helpinfo cmdhelpinfo=mprime" displays the help of MPrime.

The "ubcdcmd" is generally used to tell CPUstress which program to run after booting. "ubcdcmd=menu" means it will just show the menu.

Perhaps you can read this post by Icecube, in which he had already explained it.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 5:08 pm 
Offline

Joined: Mon Feb 21, 2005 10:59 pm
Posts: 1284
I have incorporated CPUstress image v2.1expl6 and ady's suggestion of a separate menu for CPUstress. You can check it out by downloading:

http://ubcd.mirror.fusa.be/ubcd520to53b1.xdelta

(about 6MB only) and running xdelta:

xdelta3 -d -s ubcd520.iso ubcd520to53b1.xdelta ubcd53b1.iso

to apply the delta to ubcd520.iso.

I have implemented a 1-level menu currently because I am unsure what sensible defaults exists for the other programs (eg. Stress, with its large number of parameters - number of CPUs, threads, HDD I/O etc.). So I will hold off the 2nd level menu until I see a more concrete picture of the possible defaults.

The help files are both accessible from the "CPU" menu and the "CPUstress" menu.

It's good to give CPUstress its own menu 'cos the actual version of the image can be indicated independent of the versions of the apps inside.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 7:08 pm 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
Victor, I think it would be better for the cpustrs*.txt files to be located in "/ubcd/boot/cpustress/" (and only there), so all the information and help files (and all potential changes and editions) are "contained" together in the same directory (except the boot menus, as usual). In UBCD, there is no need for additional subdirectories nor for additional copies in the menus directory. Then the corresponding line in the cfg files would be for example "
F1 /ubcd/boot/cpustress/cpustrs1.txt
".

***

Note: I am just pointing out the next little issue for the sake of customizations; there shouldn't be any conflict in an official release of UBCD.
Those 2 cpustrs*.txt files are "called" from the Syslinux boot menu using F1-F2, which are also used by other help files. In theory, there is no problem using the same Fx keys for different help files, since the calls are located in different Syslinux cfg files that are loaded independently. So an official release of UBCD should be fine with this. But some customizations could have some inconveniences. Using other keys (F3-F4) would be a possible solution for them.

***

Off-topic:
Victor,
I thought the next release (whenever it may get out) of UBCD would be V.5.2.1 (not 5.3.beta), eventually with small potential corrections and with potential feedback regarding fdubcd superfloppy image.

Also, generally speaking, it should be helpful to post md5 checksums for the files involved in the xdelta command, just in case (I had no problems, but who knows when a bad download could strike, or a mix up with the adequate original ubcd520.iso image).


Thank you and Best Regards,
Ady.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 7:20 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
Victor Chew wrote:
II have implemented a 1-level menu currently because I am unsure what sensible defaults exists for the other programs (eg. Stress, with its large number of parameters - number of CPUs, threads, HDD I/O etc.). So I will hold off the 2nd level menu until I see a more concrete picture of the possible defaults.

The help files are both accessible from the "CPU" menu and the "CPUstress" menu.

It's good to give CPUstress its own menu 'cos the actual version of the image can be indicated independent of the versions of the apps inside.


My humble opinion. I don't like that additional level of the menu. It will just add confusion.

I suggest another way:
Image

Note the differences:
  • The CPUstress image become an entry that boots into its own `menu`.
  • It's in the same level as other stress-test programs, so the user won't need to "crawl" the menus in order to find the program he/she needs.
  • The old version Mersenne Prime Test 23.9.2 is included in the menu.

You can download and try my customized cpu.cfg with this proposal:
http://dl.dropbox.com/u/70170658/ubcd/custom/cpu.cfg

Also, I'll need to rewrite the help files for this proposed change. The help file will mention what programs are included in CPUstress as well as what cheat codes are available.

EDIT: It's ready. See below post.


Last edited by Explorer09 on Wed Feb 27, 2013 6:09 pm, edited 2 times in total.

Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 7:23 pm 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
Explorer09 wrote:
Some programs, such as Stress, can have many many options to test.
Your idea can work if there's only a few options (or presets) to use. However, none of the program developers offer me such presets.

I don't see the conflict, but that's fine. Each user should better learn how to use these programs by himself anyway.
Quote:
Perhaps you can read this post by Icecube, in which he had already explained it.


Sure, I can read the post :). My intention, possible incorrectly expressed, was that the help files for CPUStress to be included in UBCD could be more clear regarding these options. Just as an example, what would be the difference between "ubcdcmd=helpinfo" alone and "ubcdcmd=helpinfo cmdhelpinfo=helpinfo"? I don't mean that I need the answer here, but that maybe the help files can be improved.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 7:29 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
ady wrote:
Victor, I think it would be better for the cpustrs*.txt files to be located in "/ubcd/boot/cpustress/" (and only there), so all the information and help files (and all potential changes and editions) are "contained" together in the same directory (except the boot menus, as usual). In UBCD, there is no need for additional subdirectories nor for additional copies in the menus directory. Then the corresponding line in the cfg files would be for example "
F1 /ubcd/boot/cpustress/cpustrs1.txt
".


Thank you. But I will rather put it in /ubcd/boot/cpustress/help . Because there are too many documents in /ubcd/boot/cpustress/ directory.

Thus the cfg line shall be "
F1 /ubcd/boot/cpustress/help/cpustrs1.txt
"

ady wrote:
My intention, possible incorrectly expressed, was that the help files for CPUStress to be included in UBCD could be more clear regarding these options. Just as an example, what would be the difference between "ubcdcmd=helpinfo" alone and "ubcdcmd=helpinfo cmdhelpinfo=helpinfo"? I don't mean that I need the answer here, but that maybe the help files can be improved.


I think the document in helpinfo is explained well enough. I leave out the cmdhelpinfo thing in cpustrs1.txt because it's too advanced and unnecessary.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 8:22 pm 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
Explorer09 wrote:
I will rather put it in /ubcd/boot/cpustress/help . Because there are too many documents in /ubcd/boot/cpustress/ directory.


I don't see "too many" files in "/ubcd/boot/cpustress/". There are pdf files in the images directory, which contains many more files. The menus have many files and they are not a problem.

IMO, the package itself is fine as it is, but UBCD can have those 2 files in "/ubcd/boot/cpustress/" without any incovenience.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Tue Feb 26, 2013 10:25 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
ady wrote:
I don't see "too many" files in "/ubcd/boot/cpustress/". There are pdf files in the images directory, which contains many more files. The menus have many files and they are not a problem.


It may be just you. For me, there is a readme.txt already, and a changes.txt. I just don't want the directory to look messy when I pack it. If I follow your suggestion, then there will be 6 txt files in the first directory in my 7z archive.

Like this:
Code:
cpustress-2.1explN/
    build.tar.xz
    changes.txt
    cpustrs1.txt
    cpustrs2.txt
    gpl-2.0.txt
    help/
        cpuburn.txt
        cpuinfo.txt
        mprime.txt
        stress.txt
    initrd.gz
    old-readme.txt
    readme.txt


By the way, I have revised the help files. Download here:
http://dl.dropbox.com/u/70170658/ubcd/c ... ustrs1.txt
http://dl.dropbox.com/u/70170658/ubcd/c ... ustrs2.txt


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Wed Feb 27, 2013 12:31 am 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
The expression "too many" files is indeed subjective. When comparing that amount of files to other directories in UBCD, it doesn't seem "too many".

As I said, the package looks fine, but UBCD doesn't necessarily have to follow the exact same structure. It could be useful for certain purposes, but doesn't have to.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Wed Feb 27, 2013 7:56 pm 
Offline

Joined: Mon Feb 21, 2005 10:59 pm
Posts: 1284
Here are the changes I have adopted:

- Image is now called 5.2b4. So the sequence is 5.2b1 => 5.2b2 => 5.2b3 => 5.2.0 => 5.2b4 => 5.2.1 etc. Reason why I am naming it this way is I am trying to keep the length to 5 characters to avoid adjusting the spaces in the title line in default.cfg in between releases. I know, I am lazy. :D Also, I don't expect a lot of widely-distributed beta releases for minor updates.

- F1, F2 now points to /ubcd/boot/cpustress/help/syslinux/cpustrs*.txt, which is where the files reside in the original package uploaded by Explorer09. I have updated those files to the latest.

- Added separate entry for CPUstress main menu as suggested by Explorer09. It is not the first entry as per your sample cfg file, but the 3rd, because I have a separate script that sorts all the entries in the cfg files alphabetically, and making exception for this one entry will require substantial changes.

Please download the xdelta file from:

http://ubcd.mirror.fusa.be/ubcd520to52b4.xdelta
md5sum: 836421e2c780ed4bead9716b4d1fa139

to verify what I have updated.

Thanks!


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Wed Feb 27, 2013 8:50 pm 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
Just in case someone has the wrong initial file (which wouldn't be a surprise):

Code:
ubcd520.iso
f359cb6d0f383c4d00903f63cfe42830
435'892'224 bytes

ubcd520to52b4.xdelta
836421e2c780ed4bead9716b4d1fa139
6'862'254 bytes

xdelta3 -d -s ubcd520.iso ubcd520to52b4.xdelta ubcd52b4.iso

ubcd52b4.iso
40b6b73449d6463af7f15dd7fc54f44b
431'661'056 bytes


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Wed Feb 27, 2013 9:50 pm 
Offline

Joined: Mon Apr 30, 2007 11:06 am
Posts: 296
Location: FL
When I saw the b in the file name I thought maybe a new beta. You could use an "R" instead of the "B" in the naming. That would be a revision or release instead of looking like beta.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Wed Feb 27, 2013 10:38 pm 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762
In ubcd/boot/cpustress/, can the old-readme.txt and the readme.txt be merged into one file?

Is the "/ubcd/boot/cpustress/bzImage" file actually used in UBCD? I would like to know its utility. Is there any situation where it could be useful (for example, can it be used if for some reason pmagic's bzImage can't? Or in such case this bzImage would not boot either?). To be clear, I am only talking about its inclusion and utility in UBCD. Its inclusion and utility in Explorer's package is clear.


@Explorer, just for convenience, in your text files you might want to express the boot lines (examples and help) in a slightly different way. For example, from you current:
Code:
     COM32  linux.c32
     APPEND /pmagic/bzImage initrd=/ubcd/boot/cpustress/initrd.gz
            root=/dev/ram0 ramdisk_size=12000 noapic ubcdcmd=cpuburn days=10

to something like:
Code:
     COM32 linux.c32 /pmagic/bzImage
     INITRD /ubcd/boot/cpustress/initrd.gz
     APPEND root=/dev/ram0 ramdisk_size=12000 noapic ubcdcmd=cpuburn days=10

so the lines are shorter and easier to read. Both formats are equivalent.

***

In addition (only partly related to CPUStress in UBCD)...

In "
ubcd/menus/grub4dos/
", the files
others.lst
memory.lst
cpu.lst

have change because of changes made to their corresponding cfg files. Apparently, the "append -" in the cfg files had a certain effect when parsed with the syslinux2grub4dos "sed" scripts :?: .


The cpustress.lst file should be deleted (leftover from previous beta).

Also custom.lst seems incorrectly affected.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Thu Feb 28, 2013 4:54 am 
Offline

Joined: Mon Nov 19, 2012 11:18 pm
Posts: 34
ady wrote:
Apparently, the "append -" in the cfg files had a certain effect when parsed with the syslinux2grub4dos "sed" scripts :?: .


I think in these lines of syslinux2grub4dos.sed you can see it (182 -194):

Quote:
# Rearrange all entries with 'LINUX' + 'APPEND -'
#
# Change:
# LINUX /ubcd/boot/memtestp/memtestp
# APPEND -
# To:
# kernel /ubcd/boot/memtestp/memtestp
# map --unmap=0:0xff

/^LINUX/ {
$!N
s|^LINUX \(.*\)\nAPPEND.*|kernel \1\nmap --unmap=0:0xff|
}
APPEND is used as end of a to be replaced string (s-command) and with everything that comes after it.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Thu Feb 28, 2013 5:33 pm 
Offline

Joined: Fri Jun 17, 2011 11:23 pm
Posts: 154
Victor Chew wrote:
- Image is now called 5.2b4. So the sequence is 5.2b1 => 5.2b2 => 5.2b3 => 5.2.0 => 5.2b4 => 5.2.1 etc. Reason why I am naming it this way is I am trying to keep the length to 5 characters to avoid adjusting the spaces in the title line in default.cfg in between releases. I know, I am lazy. :D Also, I don't expect a lot of widely-distributed beta releases for minor updates.
.
If I were you, I'll try the version like '5.3a1'. I think it is fine to have a sequence like 5.2.0 -> 5.3a1 -> 5.2.1 . Personal opinion.

Victor Chew wrote:
- Added separate entry for CPUstress main menu as suggested by Explorer09. It is not the first entry as per your sample cfg file, but the 3rd, because I have a separate script that sorts all the entries in the cfg files alphabetically, and making exception for this one entry will require substantial changes.

Ok.

ady wrote:
In ubcd/boot/cpustress/, can the old-readme.txt and the readme.txt be merged into one file?

The old readme file is just for historical reference. You can choose to not include it in the UBCD.

ady wrote:
Is the "/ubcd/boot/cpustress/bzImage" file actually used in UBCD? I would like to know its utility. Is there any situation where it could be useful (for example, can it be used if for some reason pmagic's bzImage can't? Or in such case this bzImage would not boot either?). To be clear, I am only talking about its inclusion and utility in UBCD. Its inclusion and utility in Explorer's package is clear.

"/ubcd/boot/cpustress/bzImage" is the old kernel that is used in Icecube's cpustress image. Yes, it is useful when pmagic's bzImage doesn't boot, although that seldom happens.

ady wrote:
@Explorer, just for convenience, in your text files you might want to express the boot lines (examples and help) in a slightly different way. For example, from you current:
Code:
     COM32  linux.c32
     APPEND /pmagic/bzImage initrd=/ubcd/boot/cpustress/initrd.gz
            root=/dev/ram0 ramdisk_size=12000 noapic ubcdcmd=cpuburn days=10

to something like:
Code:
     COM32 linux.c32 /pmagic/bzImage
     INITRD /ubcd/boot/cpustress/initrd.gz
     APPEND root=/dev/ram0 ramdisk_size=12000 noapic ubcdcmd=cpuburn days=10

so the lines are shorter and easier to read. Both formats are equivalent.

If I do that, you'll need to edit the syslinux2grub4dos scripts. Would you do that?


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Thu Feb 28, 2013 8:03 pm 
Offline

Joined: Sat May 08, 2010 5:26 am
Posts: 762

***
Off-topic:
Explorer09 wrote:
If I were you, I'll try the version like '5.3a1'. I think it is fine to have a sequence like 5.2.0 -> 5.3a1 -> 5.2.1 . Personal opinion.

I don't know about others, but using "5.3" just to go back to "5.2" confuses me (not to mention that the current changes only "deserve" a future new 5.2.1 at most, not a 5.3 version). Additionally, "5.3a1" would be interpreted by many (including myself) as the first alpha of a future release "5.3". OTOH, what Victor proposed uniquely identifies each release (whether final or for testing). Under the "5.2" branch (or "line"), "5.2bn" is a testing version, and once in a while a new minor (non-testing) release gets out as "5.2.m". Note that this is different than using "5.2.mbn".

In other words, what Victor proposed is adequate IMO.

Now back on topic.
***


Quote:
ady wrote:
In ubcd/boot/cpustress/, can the old-readme.txt and the readme.txt be merged into one file?

The old readme file is just for historical reference. You can choose to not include it in the UBCD.


That's Victor's decision (but generally speaking, I would be glad if some known-as unnecessary files and unnecessary directories wouldn't be included in UBCD).

Is there anything in the old readme that is currently relevant and that is not part of other included text files? For a hypothetical example, if the only relevant info would be adequate credits, that should be easily merged, shouldn't it?

Explorer09 wrote:
"/ubcd/boot/cpustress/bzImage" is the old kernel that is used in Icecube's cpustress image. Yes, it is useful when pmagic's bzImage doesn't boot, although that seldom happens.


OK. Is there any kind of hardware requirement that would be different between the two bzImage files that would affect CPUStress usage? For example, is it possible that an old system ("old" CPU, relatively small amount of RAM, yet CPU Stress programs are still able to run on such hardware) wouldn't boot with pmagic's bzImage but it would successfully boot with "
/ubcd/boot/cpustress/bzImage
"? (Note: I wouldn't recommend running CPUStress on old hardware anyway, but that's not my concern regarding this question).

My focus regarding this question is about different hardware. I remember users reporting (in the past, not recently) that they were able to use CPUStress with pmagic's bzImage but not with the original, and other users stating exactly the opposite, depending on the hardware they were trying to test.

If both bzImage files are really useful and really necessary, then maybe this should be mentioned somewhere. If only one bzImage is needed to cover all practical cases, then the unnecessary file should be only part of the package (for testing purposes for example) but not part of UBCD.

Quote:
ady wrote:
@Explorer, just for convenience, in your text files you might want to express the boot lines (examples and help) in a slightly different way. For example, from you current:
Code:
     COM32  linux.c32
     APPEND /pmagic/bzImage initrd=/ubcd/boot/cpustress/initrd.gz
            root=/dev/ram0 ramdisk_size=12000 noapic ubcdcmd=cpuburn days=10

to something like:
Code:
     COM32 linux.c32 /pmagic/bzImage
     INITRD /ubcd/boot/cpustress/initrd.gz
     APPEND root=/dev/ram0 ramdisk_size=12000 noapic ubcdcmd=cpuburn days=10

so the lines are shorter and easier to read. Both formats are equivalent.

If I do that, you'll need to edit the syslinux2grub4dos scripts. Would you do that?


I don't think there would be a need to change syslinux2grub2dos scripts, since Victor already uses a similar format in UBCD as I am suggesting here for your help files (it is not a direct copy from your help files). My suggestion was (is) for convenience and easier reading in your help files (there wouldn't be a need to clarify the artificial use of an extra line for 'append'). But, if the scripts are a concern regarding these lines, I am willing to test and report back if necessary.


Top
 Profile  
 
 Post subject: Re: Explorer's modification of CPUstress image
PostPosted: Thu Feb 28, 2013 10:20 pm 
Offline

Joined: Mon Feb 21, 2005 10:59 pm
Posts: 1284
Quote:
When I saw the b in the file name I thought maybe a new beta. You could use an "R" instead of the "B" in the naming. That would be a revision or release instead of looking like beta.
I prefer "b" really. Technically it is a beta and not for general consumption. "r" gives the impression that it is an official revision.

Quote:
Is the "/ubcd/boot/cpustress/bzImage" file actually used in UBCD?
I have actually removed it, seeing that we are not using it, and Explorer09 has not included it in his archive.

Quote:
Just for convenience, in your text files you might want to express the boot lines (examples and help) in a slightly different way
I have reformatted it to fit the style used by the other entries in the UBCD menus.

Quote:
Apparently, the "append -" in the cfg files had a certain effect when parsed with the syslinux2grub4dos "sed" scripts :?: .
I have updated syslinux2grub4dos scripts to deal with this.

Quote:
The cpustress.lst file should be deleted (leftover from previous beta).
Deleted.

Quote:
Also custom.lst seems incorrectly affected.
Fixed.

You can download the xdelta file for 5.2b5 here:

http://ubcd.mirror.fusa.be/ubcd520to52b5.xdelta
md5sum: 1364f4395624b8e3a6825b7b178c92fc

to verify what I have updated.


Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 60 posts ]  Go to page 1, 2, 3  Next

All times are UTC - 8 hours


Who is online

Users browsing this forum: No registered users and 4 guests


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  

Protected by Anti-Spam ACP Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group