Currently I'm developing what will eventually be a biggish suite of test programs to help in determining potential problems in the use of C static analysis tools with different C compilers. To that end I've recently installed clang and tcc to complement the gcc and g++ that I already had on the PC.
The PC configuration is:
IBM ThinkPad R40e
CPU 2.4 MHz Celeron M
1 GB RAM
36 GB HDD
running tahrpup 6.0.5
When I installed both tcc and clang, it took ages for PPM to do the post-download installation processing (making total install time just over 30 mins for clang) and the processor utilisation shot up to over 80% with CPU temperature over 70 degrees (F or C?) most of the time.
What is it doing that creates this huge CPU load?
CPU load when installing s/w packages with PPM
-
- Posts: 276
- Joined: Thu 29 Mar 2018, 22:13
- Location: West Wales bandit country
CPU load when installing s/w packages with PPM
Sometimes I post mindfully, sometimes not mindfully, and sometimes both mindfully and not mindfully. It all depends on whether and when my mind goes walkies while I'm posting :?
-
- Posts: 276
- Joined: Thu 29 Mar 2018, 22:13
- Location: West Wales bandit country
That would make some sense, as the R40e I'm using for it has a rather slow HDD. Thing is, though, that the HD activity light at the back of the keyboard does not appear to show much use of the HD - though this might, I assume, be accounted for by relatively infrequent disk accesses for moderate-sized chunks of info that are built up and checked in RAM.mikeb wrote:Hmm from memory manically dependency checking is a major element which includes checking existing binaries and libraries are not adversely affected....note I have not dived into it for a long time and install manually using xarchive or use sfs so sorry for the vagueness.
mike
Similar installs on a netbook:
HP Mini 110-1110
dual-core Intel Atom N270n @ 1.6 GHz
2 GB RAM
120 GB SSD
tahrpup 6.0.5 PAE
do go faster but still generate a high CPU load.
Sometimes I post mindfully, sometimes not mindfully, and sometimes both mindfully and not mindfully. It all depends on whether and when my mind goes walkies while I'm posting :?
FWIW, I built a clang squashfs module in a Stretch-Live environment. It only took a few minutes and passed the Hello World test.When I installed both tcc and clang, it took ages for PPM to do the post-download installation processing (making total install time just over 30 mins for clang) and the processor utilisation shot up to over 80% with CPU temperature over 70 degrees (F or C?) most of the time.
Using the Puppy PPM will never work as well as installing packages in their native environment.
-
- Posts: 1543
- Joined: Mon 22 Feb 2016, 19:43
Quite sure mikeb is right. There's a way to turn off the "manically dependency checking" but only in really new woof-CE builds.
- Attachments
-
- Screenshot.png
- (41.19 KiB) Downloaded 137 times
-
- Posts: 276
- Joined: Thu 29 Mar 2018, 22:13
- Location: West Wales bandit country
On thinking about this, I have another, perhaps more important question:mikeb wrote:Hmm from memory manically dependency checking is a major element which includes checking existing binaries and libraries are not adversely affected....note I have not dived into it for a long time and install manually using xarchive or use sfs so sorry for the vagueness.
I use gcc-4.8 most of the time but also have gcc-4.6 installed. If the install order is 4.6 then 4.8, am I running exactly the same 4.8 as if I had installed in the order 4.8 then 4.6?
If the configurations differ, then this has some pretty fundamental implications for how I design the test suite that I'm currently working on.
Sometimes I post mindfully, sometimes not mindfully, and sometimes both mindfully and not mindfully. It all depends on whether and when my mind goes walkies while I'm posting :?
-
- Posts: 276
- Joined: Thu 29 Mar 2018, 22:13
- Location: West Wales bandit country
Thanks. One of the jobs that my test suite will be doing is determining exactly what the installed configuration is. Now I know what to look for, I can make sure the test programs are properly designed.mikeb wrote:Last installed would normally be on top/overwrite so yes order could well make a difference.
Note puppy's sfs handling is the reverse of this.
If each gcc version gets a separate folder this may prevent such conflicts.
mike
Sometimes I post mindfully, sometimes not mindfully, and sometimes both mindfully and not mindfully. It all depends on whether and when my mind goes walkies while I'm posting :?