encounter crash when placing

[Dick Hou]

soc encounter techsite error ?

When I use SocEncounter to place the imported design,it crash every time,and tell me the internal error(SEGV),anybody have the same problem?
(soce 3.1 under rh 7.3)

 

[Dick Hou]

soc encounter terminated by internal

<CMD> amoebaPlace
Extracting standard cell pins and blockage ......
Pin and blockage extraction finished
Extracting macro/IO cell pins and blockage ......
Pin and blockage extraction finished
*** Starting "Amoeba(TM) placement v0.254.2.18 (mem=57.0M)" ...
Options: gp=mq-medium dp=medium
Options: congestDrivenI75 congestDrivenQ congestDrivenR1
#std cell=116 #block=0 (0 floating + 0 preplaced) #ioInst=0 #net=172 #term=510 #term/net=2.97, #fixedIo=0, #floatIo=0, #fixedPin=0, #floatPin=66
Total std cell len = 1.4953 (mm), area = 0.0254 (mm^2)
*info: identify 1 tech site => 1 tech site pattern
*info: estimated cell pwr/gnd rail width = 2.120 um
*info: average module density = 0.884
*info: density for the design = 0.884
= stdcell_area 787 (25420 mu^2) / alloc_area 890 (28747 mu^2).
*info: fixed term percentage = 12.671%
*info: fixed term center of gravity = (0 0) in coreBox (0 0, 17000 17000)
*info: partition deviation = 0.050
0|: Est. net length = 2.104e+04 (1.22e+04 8.84e+03) cpu=0:00:00.0 mem=57.0M
1-: Est. net length = 1.909e+04 (1.12e+04 7.90e+03) cpu=0:00:00.0 mem=57.0M
2|: Est. net length = 1.719e+04 (8.67e+03 8.51e+03) cpu=0:00:00.0 mem=57.0M
..+: Est. net length = 1.370e+04 (8.30e+03 5.39e+03) cpu=0:00:00.0 mem=57.0M
*** cost = 1.370e+04 (8.30e+03 5.39e+03) (cpu for global=0:00:00.1) ***
Options: gp=mq-medium dp=medium
Starting fine tune place ...

CPU Time for Phases I and II = 0:00:00.0, Real Time = 0:00:00.0
Statistics of distance of Instance movement in detailed placement
maximum (X+Y) = 94.80 um
inst (acc_r_reg[6]) with max move: (83.6, 85) -> (22.8, 51)
mean (X+Y) = 36.32 um
Total instances moved : 102
*** cpu=0:00:00.0 mem=57.0M ***
Total net length = 1.084e+04 (6.582e+03 4.255e+03) (ext = 1.636e+03)
*** End placement (cpu=0:00:00.2, real=0:00:00.0, mem=57.0M) ***
Encounter terminated by internal (SEGV) error/signal...

*** Memory Usage v0.51.2.2 (Current mem = 56.996M, initial mem = 36.270M) ***
--- Ending "First Encounter" (totcpu=0:00:10.6, real=0:04:38, mem=57.0M) ---



:?: the error occured at end placement, why?

 

[loroychen]

encounter crash partition

If LEF file only has two routing layers (metal1 & metal2), SOC encounter will crash when it want to place the cells. This is my observation. I do not know why. I have exactly same problem and I do not know how to fix.

 

[konqueror]

how to route using to metal layers in soc encounter.it uses 3 metal layers .does any body have a solution to this.

 

[zhustudio]

it is common software bugs. you can update encounter. the current version of encounter is 4.1 user3. it is much faster than old version.

 

[raj_rohit10]

hi,
In any Unix flavour there is a command "TRUSS" by waching truss window uyou can be more specific about the reason why it is failing?
Syntax: $truss ./a.out

 

[ryang]

Try to update SOC Encounter or apply the patches from cadence.