=================== BUG #1593: LATEST MODIFICATIONS ==================
http://savannah.nongnu.org/bugs/?func=detailbug&bug_id=1593&group_id=2348

Changes by: Mathieu Roy <[EMAIL PROTECTED]>
Date: 2002-Nov-15 08:23 (Europe/Paris)

------------------ Additional Follow-up Comments ----------------------------
Florian, can you give a way to easily get the system arch for any of the arch 
listed above ?

(for instance, results of uname -sm)



=================== BUG #1593: FULL BUG SNAPSHOT ===================


Submitted by: fweps                     Project: Package DataBase View          
Submitted on: 2002-Nov-03 21:33
Category:  core                         Severity:  3                            
Bug Group:  None                        Resolution:  None                       
Assigned to:  fweps                     Status:  Open                           

Summary:  CPUSPEED detection is intel-specific

Original Submission:  The way we currently determine CPUSPEED is intel-specific.
Sadly, /proc/cpuinfo varies wildly on different
architectures and sometimes doesn't even contain CPU
frequency.

Grepping for "cpu MHz" works on i386 and ia64.

I'll collect samples of /proc/cpuinfo for all
architectures I have access to, and come up with a cool
solution.

Follow-up Comments
*******************

-------------------------------------------------------
Date: 2002-Nov-15 08:23             By: yeupou
Florian, can you give a way to easily get the system arch for any of the arch 
listed above ?

(for instance, results of uname -sm)

-------------------------------------------------------
Date: 2002-Nov-04 00:41             By: yeupou
Or maybe we should get over it and set a 7s sleep for everybody ?

-------------------------------------------------------
Date: 2002-Nov-04 00:37             By: yeupou
Hum, this find problem will be very painy. 

For instance, with the itanium bipro, we got only values of processors one by 
one.

And no way to be sure that it will be portable in the future... 

I havent received any answer to my bug report to gnu findutils... 

Maybe we should use something else that find ? (as possible, usual on GNU/Linux 
or other Free OSes)



-------------------------------------------------------
Date: 2002-Nov-04 00:31             By: fweps
sparc

[EMAIL PROTECTED]:~$ cat /proc/cpuinfo
cpu             : TI UltraSparc II  (BlackBird)
fpu             : UltraSparc II integrated FPU
promlib         : Version 3 Revision 17
prom            : 3.17.0
type            : sun4u
ncpus probed    : 2
ncpus active    : 2
Cpu0Bogo        : 897.84
Cpu0ClkTck      : 000000001ad2c280
Cpu2Bogo        : 897.84
Cpu2ClkTck      : 000000001ad2c280
MMU Type        : Spitfire
State:
CPU0:           online
CPU2:           online
[EMAIL PROTECTED]:~$

----------------------------------------------------------------------

ia64

[EMAIL PROTECTED]:~$ cat /proc/cpuinfo
processor  : 0
vendor     : GenuineIntel
arch       : IA-64
family     : Itanium
model      : 0
revision   : 6
archrev    : 0
features   : standard
cpu number : 0
cpu regs   : 4
cpu MHz    : 800.178000
itc MHz    : 800.178000
BogoMIPS   : 796.91

processor  : 1
vendor     : GenuineIntel
arch       : IA-64
family     : Itanium
model      : 0
revision   : 6
archrev    : 0
features   : standard
cpu number : 0
cpu regs   : 4
cpu MHz    : 800.178000
itc MHz    : 800.178000
BogoMIPS   : 796.91

[EMAIL PROTECTED]:~$

----------------------------------------------------------------------

arm

(base)[EMAIL PROTECTED]:~$ cat /proc/cpuinfo
Processor       : Intel StrongARM-110 rev 3 (v4l)
BogoMIPS        : 185.95
Features        : swp half 26bit fastmult 

Hardware        : Rebel-NetWinder
Revision        : 52ff
Serial          : 0000000000000c5c
(base)[EMAIL PROTECTED]:~$

----------------------------------------------------------------------

alpha

[EMAIL PROTECTED]:~$ cat /proc/cpuinfo
cpu                     : Alpha
cpu model               : PCA56
cpu variation           : 0
cpu revision            : 0
cpu serial number       : Linux_is_Great!
system type             : EB164
system variation        : SX164
system revision         : 0
system serial number    : MILO-2.2-17
cycle frequency [Hz]    : 533333333 
timer frequency [Hz]    : 1024.00
page size [bytes]       : 8192
phys. address bits      : 40
max. addr. space #      : 127
BogoMIPS                : 1055.68
kernel unaligned acc    : 0 (pc=0,va=0)
user unaligned acc      : 0 (pc=0,va=0)
platform string         : N/A
cpus detected           : 0
[EMAIL PROTECTED]:~$

----------------------------------------------------------------------

hppa

[EMAIL PROTECTED]:~$ cat /proc/cpuinfo
processor       : 0
cpu family      : PA-RISC 2.0
cpu             : PA8600 (PCX-W+)
cpu MHz         : 550.000000
model           : 9000/800/A500-5X
model name      : Crescendo 550
hversion        : 0x00005d50
sversion        : 0x00000491
I-cache         : 512 KB
D-cache         : 1024 KB (WB)
ITLB entries    : 160
DTLB entries    : 160 - shared with ITLB
bogomips        : 1097.72
software id     : 1583770588

[EMAIL PROTECTED]:~$

----------------------------------------------------------------------

s390

(base)[EMAIL PROTECTED]:~$ cat /proc/cpuinfo
vendor_id       : IBM/S390
# processors    : 1
bogomips per cpu: 632.42
processor 0: version = FF,  identification = 049321,  machine = 9672
(base)[EMAIL PROTECTED]:~$

----------------------------------------------------------------------

powerpc

[EMAIL PROTECTED] cat /proc/cpuinfo
cpu             : 750CX
temperature     : 11-13 C (uncalibrated)
clock           : 499MHz
revision        : 34.20 (pvr 0008 2214)
bogomips        : 996.14
machine         : PowerBook4,1
motherboard     : PowerBook4,1 PowerBook2,2 MacRISC2 MacRISC Power Macintosh
L2 cache        : 256K unified
memory          : 384MB
pmac-generation : NewWorld
[EMAIL PROTECTED]

----------------------------------------------------------------------

i386

[EMAIL PROTECTED] cat /proc/cpuinfo
processor       : 0
vendor_id       : GenuineIntel
cpu family      : 6
model           : 3
model name      : Pentium II (Klamath)
stepping        : 3
cpu MHz         : 233.140
cache size      : 512 KB
fdiv_bug        : no
hlt_bug         : no
f00f_bug        : no
coma_bug        : no
fpu             : yes
fpu_exception   : yes
cpuid level     : 2
wp              : yes
flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge mca cmov 
mmx
bogomips        : 465.30

[EMAIL PROTECTED]

----------------------------------------------------------------------

Debian architectures missing from this list:

    * Motorola 680x0
    * MIPS
    * MIPS (DEC)



-------------------------------------------------------
Date: 2002-Nov-03 22:41             By: fedemp
Here you are some examples I have just collected:

Machine               BogoMIPS
[EMAIL PROTECTED]              699.59
Dual [EMAIL PROTECTED]        1993.77
Cuat. [EMAIL PROTECTED]      796.91
Dual Alpha [EMAIL PROTECTED]   1330.04
Dual Alpha [EMAIL PROTECTED]     998.56


-------------------------------------------------------
Date: 2002-Nov-03 22:13             By: fweps
There is a table for rough bogomips|-->clock mappings at
  http://www.clifton.nl/bogomips-2.html#ss2.2
Maybe this will become useful. 

I'll do a tour of the Debian project machines and collect all
/proc/cpuinfo listings later tonight.

Florian


-------------------------------------------------------
Date: 2002-Nov-03 22:06             By: yeupou
That's correct but I do not know how much bogomips should consider as critical 
for find.

P120 = approx 240 bogomips
P350 = approx 700 bogomips

bi-P800 = apprx 3200 bogomips
http://savannah.gnu.org/phpsysinfo/

It could help to determine that.

-------------------------------------------------------
Date: 2002-Nov-03 21:58             By: fedemp
Hi, I wander if the 'bogomips' line is more "portable".
If it is, the bogomips counter could be used as an
approximation of the computational power instead of the
Mhz.


-------------------------------------------------------
Date: 2002-Nov-03 21:54             By: yeupou
Ok. 

If GNU find wasnt buggy, this part would be absolutely usefull... What a pain. 

Currently, if we do not deal with linux kernel, we skip the CPUSPEED and 
consider that we need to do the 7s sleep.
We should maybe test also whether we deal with an arch. we know how to get the 
CPU frequency or not (in this case, the test would be also skipped).




CC list is empty


No files currently attached


For detailed info, follow this link:
http://savannah.nongnu.org/bugs/?func=detailbug&bug_id=1593&group_id=2348


Reply via email to