Lm-sensors is nothing useful when you use it alone. The hardware monitoring info is available already in BIOS. And I don't know what temperatures people would want to see when he just boots up his computer?Victor Chew wrote:I have updated to CPUStress V2.3.1 as per your suggestions.
I didn't find an entry for Sensors, so I put one together.
Please feel free to make corrections as you see fit.
Explorer's modification of CPUstress image
Moderators: Icecube, StopSpazzing
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
-
- Posts: 1368
- Joined: Mon Feb 21, 2005 10:59 pm
- Contact:
Re: Explorer's modification of CPUstress image
I see. So I will remove the entry.
Thanks!
Thanks!
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
I make a correction to my last post.
Some laptops do not show hardware monitoring info in the BIOS. I'm not sure about EFI computers, though.
My second statements still stands. Hardware monitoring isn't very useful when your CPU is running only monitoring tool and nothing else.
Some laptops do not show hardware monitoring info in the BIOS. I'm not sure about EFI computers, though.
My second statements still stands. Hardware monitoring isn't very useful when your CPU is running only monitoring tool and nothing else.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.3.2 (3 May 2013)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.2.7z
Size: 5914515 bytes
MD5: c0cc64ed61f32305b2afe42ee8041799
SHA1: 565c925fe471b5ee97b2b278a079496f372dee31
https://dl.dropboxusercontent.com/u/701 ... s-2.3.2.7z
Size: 5914515 bytes
MD5: c0cc64ed61f32305b2afe42ee8041799
SHA1: 565c925fe471b5ee97b2b278a079496f372dee31
- Updated kernel to 3.8.11.
- Add clarifications to licenses. (nonfree.txt)
- Minor fixes in scripts and documentation.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.3.3 (19 May 2013)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.3.7z
Size: 5919410 bytes
MD5: 8fdf6764b954fcb9993692cd0532f160
SHA1: 5781a8f67b32119b0095b37a547f176846390cd1
(UPDATE: 24 May 2013.
The previous release have mixed line endings in the documents. This is a minor problem but is corrected today, so please re-download and use the new hashes below.
Size: 5919335 bytes
MD5: eec779cbd8bc8d3e7d5f3188a86c8a31
SHA1: 2e9da87b6157fe86758d047b0e342e3efb6ad38e)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.3.7z
Size: 5919410 bytes
MD5: 8fdf6764b954fcb9993692cd0532f160
SHA1: 5781a8f67b32119b0095b37a547f176846390cd1
(UPDATE: 24 May 2013.
The previous release have mixed line endings in the documents. This is a minor problem but is corrected today, so please re-download and use the new hashes below.
Size: 5919335 bytes
MD5: eec779cbd8bc8d3e7d5f3188a86c8a31
SHA1: 2e9da87b6157fe86758d047b0e342e3efb6ad38e)
- Updated kernel to 3.8.13.
- Rebuilt BusyBox and eglibc (same version).
- The "ubcdargs" boot option is now also shown in "UBCD args" field in splash screen.
- Add "help/linpack.txt".
- Minor fixes in scripts and documentation.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.3.4 (18 June 2013)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.4.7z
Size: 5920077 bytes
MD5: 1749e4051209cbd9e210fee19ffcd4ab
SHA1: b9d40f772ca9be3ce357fee0fc3a28b6532634d4
https://dl.dropboxusercontent.com/u/701 ... s-2.3.4.7z
Size: 5920077 bytes
MD5: 1749e4051209cbd9e210fee19ffcd4ab
SHA1: b9d40f772ca9be3ce357fee0fc3a28b6532634d4
- Updated Sensors to v3.3.4 (from v3.3.3).
- Updated eglibc to 2.17-r23298 (from 2.17-r22665).
- Rebuilt systester-cli (same version).
- Added sensors-detect wrapper.
- Minor fixes in scripts and documentation.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.3.5 (20 July 2013)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.5.7z
Size: 5829201 bytes
MD5: 89a46e05e4827c3f806dd2d4bca05026
SHA1: d245eb99be864b3b4434ba033fce0106b0c4798a
There are a few things you need to do when you incorporate this CPUstress image to UBCD:
1. I've updated Intel LINPACK to v11.0.5, so please change the version number in the Isolinux menu entry and the entry on website.
2. The Isolinux entry of System Stability Tester has the description saying it "Can create up to 32 threads." The actual limit is 64 threads, so please change the text. Thanks. (The 32 threads limit is artificially made in GUI and doesn't apply to me. I've already reported the issue.)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.5.7z
Size: 5829201 bytes
MD5: 89a46e05e4827c3f806dd2d4bca05026
SHA1: d245eb99be864b3b4434ba033fce0106b0c4798a
- Updated Intel LINPACK to v11.0.5 (from v11.0.3)
- stresscpu2 is rebuilt with dynamic linking.
- Patched and rebuilt systester-cli. See the compile note for the list of changes (there are 10 changes).
- Updated kernel to 3.10.1. Added drivers for keypads.
- Rebuilt eglibc (2.17-r23365).
- Updated BusyBox to v1.21.1 (from v1.21.0)
There are a few things you need to do when you incorporate this CPUstress image to UBCD:
1. I've updated Intel LINPACK to v11.0.5, so please change the version number in the Isolinux menu entry and the entry on website.
2. The Isolinux entry of System Stability Tester has the description saying it "Can create up to 32 threads." The actual limit is 64 threads, so please change the text. Thanks. (The 32 threads limit is artificially made in GUI and doesn't apply to me. I've already reported the issue.)
-
- Posts: 1368
- Joined: Mon Feb 21, 2005 10:59 pm
- Contact:
Re: Explorer's modification of CPUstress image
Got it!@Victor
There are a few things you need to do when you incorporate this CPUstress image to UBCD:
1. I've updated Intel LINPACK to v11.0.5, so please change the version number in the Isolinux menu entry and the entry on website.
2. The Isolinux entry of System Stability Tester has the description saying it "Can create up to 32 threads." The actual limit is 64 threads, so please change the text. Thanks. (The 32 threads limit is artificially made in GUI and doesn't apply to me. I've already reported the issue.)
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
I have an updated kernel for CPUstress:
https://dl.dropboxusercontent.com/u/701 ... 3.10.5.tar
MD5: 123354d6567a0288b090e4687c5465ac
SHA1: e0455c0165d764d82ea90ffd39be8edf042bce14
Because this is only a kernel update, I don't want to make a new CPUstress release just for it.
https://dl.dropboxusercontent.com/u/701 ... 3.10.5.tar
MD5: 123354d6567a0288b090e4687c5465ac
SHA1: e0455c0165d764d82ea90ffd39be8edf042bce14
Because this is only a kernel update, I don't want to make a new CPUstress release just for it.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
Kernel 3.10.7-cpustress
https://dl.dropboxusercontent.com/u/701 ... 3.10.7.tar
MD5: eebfca459da78cc9fb950223065226ad
SHA1: 283c279da56cae6641764c21a6993fc5f7ea5613
https://dl.dropboxusercontent.com/u/701 ... 3.10.7.tar
MD5: eebfca459da78cc9fb950223065226ad
SHA1: 283c279da56cae6641764c21a6993fc5f7ea5613
-
- Posts: 1368
- Joined: Mon Feb 21, 2005 10:59 pm
- Contact:
Re: Explorer's modification of CPUstress image
Got it. Thanks!
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.3.6 (15 September 2013)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.6.7z
Size: 6181222 bytes
MD5: 331f931144068ab6c54c54dc7691b543
SHA1: 20ee26893fb7f140c6324f4509a4147877c84336
It seems that you forgot what I said when you release UBCD 5.2.6, so I repeat again:
1. I've updated Intel LINPACK to v11.1.0, so please change the version number in the Isolinux menu entry and the entry on website.
2. The Isolinux entry of System Stability Tester has the description saying it "Can create up to 32 threads." The actual limit is 64 threads, so please change the text.
Update: (2013-10-24)
Kernel 3.11.6-cpustress
https://dl.dropboxusercontent.com/u/701 ... 3.11.6.tar
MD5: 2c814968c948a1e8b8b4aaf1c9e8fd19
SHA1: 4b94b3d2a658bb155080afb6e7ad5db2071d7450
https://dl.dropboxusercontent.com/u/701 ... s-2.3.6.7z
Size: 6181222 bytes
MD5: 331f931144068ab6c54c54dc7691b543
SHA1: 20ee26893fb7f140c6324f4509a4147877c84336
- Updated Intel LINPACK to v11.1.0 (from v11.0.5)
- Updated kernel to 3.11.1. Changes in this kernel build:
- Disable CC_OPTIMIZE_FOR_SIZE
- Support x86 extended platforms (Intel MID, Medfield)
- Add hwmon drivers connected through SPI and IPMI
- Removed hwmon drivers that depend on Industrial I/O
- Removed I2C system bus drivers (for embedded)
- Updated eglibc to 2.18-r23929 (from 2.17-r23365). Also I removed several functions in this libc build, so the library is slightly smaller. See the compile note for details.
- Improvements to the build script.
- Minor fixes in scripts and documentation.
It seems that you forgot what I said when you release UBCD 5.2.6, so I repeat again:
1. I've updated Intel LINPACK to v11.1.0, so please change the version number in the Isolinux menu entry and the entry on website.
2. The Isolinux entry of System Stability Tester has the description saying it "Can create up to 32 threads." The actual limit is 64 threads, so please change the text.
Update: (2013-10-24)
Kernel 3.11.6-cpustress
https://dl.dropboxusercontent.com/u/701 ... 3.11.6.tar
MD5: 2c814968c948a1e8b8b4aaf1c9e8fd19
SHA1: 4b94b3d2a658bb155080afb6e7ad5db2071d7450
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.3.7 (22 November 2013)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.7.7z
Size: 6237248 bytes
MD5: 0df0a37e6d34fff3b302a2108585ca77
SHA1: b4dfb78e90f389b03769d8f50fe06948a5d661d4
Kernel 3.12.5-cpustress
https://dl.dropboxusercontent.com/u/701 ... 3.12.5.tar
MD5: 2c09927e63040b17b621eef7aaef5f87
SHA1: 468d214eb7a8cb690c91368f60cac385b7eed428
(Older kernel builds are available: 3.11.10, 3.10.24)
https://dl.dropboxusercontent.com/u/701 ... s-2.3.7.7z
Size: 6237248 bytes
MD5: 0df0a37e6d34fff3b302a2108585ca77
SHA1: b4dfb78e90f389b03769d8f50fe06948a5d661d4
- Updated kernel to 3.12.1.
- Rebuilt eglibc (2.18-r24307).
- Updated Intel LINPACK to v11.1.1 (from v11.1.0)
- Patched and rebuilt stresscpu2 to accept '--help' option on the command-line.
- Rebuilt systester-cli with GMP 5.1.3.
- Minor fixes in scripts and documentation.
Kernel 3.12.5-cpustress
https://dl.dropboxusercontent.com/u/701 ... 3.12.5.tar
MD5: 2c09927e63040b17b621eef7aaef5f87
SHA1: 468d214eb7a8cb690c91368f60cac385b7eed428
(Older kernel builds are available: 3.11.10, 3.10.24)
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
CPUstress 2.4.0 (24 February 2014)
https://dl.dropboxusercontent.com/u/701 ... s-2.4.0.7z
Size: 4607939 bytes
MD5: 06c61c629f327866eaa6842d6d4d5aa7
SHA1: f2daa8ca164c232d09eefc6ac5a6c4c79c95091a
Programs updated or rebuilt:
1. I changed the compression of initrd from Gzip to XZ, so the isolinux boot entries need to be changed as well.
From this:
https://dl.dropboxusercontent.com/u/701 ... s-2.4.0.7z
Size: 4607939 bytes
MD5: 06c61c629f327866eaa6842d6d4d5aa7
SHA1: f2daa8ca164c232d09eefc6ac5a6c4c79c95091a
Programs updated or rebuilt:
- Updated Intel LINPACK to v11.1.2 (from v11.1.1)
- Updated Sensors to v3.3.5 (from v3.3.4)
- Updated BusyBox to v1.22.1 (from v1.21.1)
- Updated eglibc to 2.19-r25243 (from 2.18-r24307)
- Updated kernel to 3.13.5. Config changes in this kernel build:
- No more CONFIG_IKCONFIG
- Block layer and ext2 filesystem support removed.
- Timer interrupt frequency is set to 100Hz (from 250Hz)
- Remove drivers of mouse and HID devices that are useless with cpustress.
- Changed kernel and initrd compression from gzip to XZ. Build scripts and help documents are all updated to reflect this change. This makes the compressed kernel+initrd image 648KiB smaller, although the memory required for decompression increases from 600KiB to 800KiB.
- Use --ultra-brute switch in upx to re-compress linpack and mprime binaries. Also compress sensors, stress, stresscpu2, and systester-cli binaries.
1. I changed the compression of initrd from Gzip to XZ, so the isolinux boot entries need to be changed as well.
From this:
To this:COM32 linux.c32 /ubcd/boot/cpustress/bzImage
INITRD /ubcd/boot/cpustress/initrd.gz
2. Intel Optimized LINPACK Benchmark here has been updated to version 11.1.2. You probably need to correct the version number in the isolinux boot entries.COM32 linux.c32 /ubcd/boot/cpustress/bzImage
INITRD /ubcd/boot/cpustress/initrd.xz
-
- Posts: 1368
- Joined: Mon Feb 21, 2005 10:59 pm
- Contact:
Re: Explorer's modification of CPUstress image
Got it. Thanks!
Re: Explorer's modification of CPUstress image
Please forgive my ignorance.
I am curious about:
And regarding:
(Note: Stress tests are sometimes used to check older systems - which may or may not be advisable, but that's a different matter.)
TIA,
Ady.
I am curious about:
Are those not relevant (anymore)?Explorer09 wrote: Block layer and ext2 filesystem support removed.
And regarding:
Does this change affect older hardware in any significant way?Explorer09 wrote: Changed kernel and initrd compression from gzip to XZ.
(Note: Stress tests are sometimes used to check older systems - which may or may not be advisable, but that's a different matter.)
TIA,
Ady.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
The ext2 support was needed in the previous initrd format. Since I have switched to initramfs (compressed cpio) format, they are not relevant anymore.ady wrote:Are those not relevant (anymore)?
(The entire image will be run on memory. I have no intention to support hard disk access in cpustress.)
It will increase the decompression memory a little bit. Although I'm very careful about this and choose a small dictionary size for compressing initrd. (The decompression of initrd is the only thing that needs additional memory overhead. Others such as kernel bzImage and UPX executables are decompressed in-place.) Of course the decompression will be slower, but consider the read speed of the CDROM (when loading the kernel and initrd images), I don't think it is relevant, although I have no old hardware to test it.ady wrote: Does this change affect older hardware in any significant way?
Re: Explorer's modification of CPUstress image
@Explorer,
Thank you for the prompt answer
.
Thank you for the prompt answer

-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
Hello everyone.
I finally have the time to finish this new version of the image.
CPUstress 2.5.0 (15 August 2014)
https://dl.dropboxusercontent.com/u/701 ... s-2.5.0.7z
Size: 4694727 bytes
MD5: a05bdc5d79bad6ef5cbb5e26d4f32287
Programs updated or rebuilt:
Also, MPrime 28.5 is not ready to be included yet. The new mprime binary now links to a shared libstdc++, and I'll need some time to re-compile it so that libstdc++ can be linked statically.
EDIT: The new version is ready. Use 2.5.1 instead (see below post).
I finally have the time to finish this new version of the image.
CPUstress 2.5.0 (15 August 2014)
https://dl.dropboxusercontent.com/u/701 ... s-2.5.0.7z
Size: 4694727 bytes
MD5: a05bdc5d79bad6ef5cbb5e26d4f32287
Programs updated or rebuilt:
- Updated Intel LINPACK to v11.1.3 (from v11.1.2)
- Rebuilt systester-cli with GMP 6.0.0. Also GMP is now built as a fat binary.
- Rebuilt BusyBox and added 'mv' command.
- Updated libgcc to v4.9.1 (from v4.7.3)
- Updated kernel to 3.16.1.
- All programs and libraries (except cpuburn-in, linpack, and mprime) are re-compiled with a custom toolchain so that they are compatible with i486 CPUs. All CPUstress releases since v2.1expl3 and until this version were not really compatible with i486 and Pentium CPUs, due to the Ubuntu GCC that generates i686-only code. This regression is now fixed.
- New wrapper script for Mersenne Prime test (mprime). This wrapper can now ask user for which type of torture test to run, and tune the torture test settings. Help document for mprime is also re-written.
- Various optimizations in shell scripts.
Also, MPrime 28.5 is not ready to be included yet. The new mprime binary now links to a shared libstdc++, and I'll need some time to re-compile it so that libstdc++ can be linked statically.
EDIT: The new version is ready. Use 2.5.1 instead (see below post).
Last edited by Explorer09 on Wed Aug 20, 2014 2:01 am, edited 2 times in total.
-
- Posts: 178
- Joined: Fri Jun 17, 2011 11:23 pm
Re: Explorer's modification of CPUstress image
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
https://dl.dropboxusercontent.com/u/701 ... s-2.5.1.7z
Size: 4626848 bytes
MD5: cd70fdcc52ff906399e3944909042995
SHA1: 962af04a85e43d5646db86687718c1fde784cdb9
- Updated Mersenne Prime test to v28.5 build 2 (from v27.9 build 1)
- Minor fixes in scripts and documentation.