Skip to content

Failing config files #6

Description

@agyaglikci

Most of the config files in the repo lead to either a segmentation fault or a floating point error. I basically clone, compile, and run cacti w/o touching anything. I use a Ubuntu 16.04.3 LTS system with kernel 4.15.0-24-generic. Could you take a look at that?

You can find a summary of these config files below:

This is what I do:

git clone https://github.com/HewlettPackard/cacti.git Cacti7
cd Cacti7
make
./cacti -infile sample_config_files/lpddr3_cache.cfg  

Here are the errors I get with the provided config files:

Config File Behavior
cache.cfg Finishes successfully
ddr3.cfg Throws Segmentation fault (core dumped)
dram.cfg Throws Floating point exception (core dumped)
lpddr.cfg Throws Segmentation fault (core dumped)
2DDRAM_micron1Gb.cfg Throws Floating point exception (core dumped)
2DDRAM_Samsung2GbDDR2.cfg Throws Floating point exception (core dumped)
3DDRAM_Samsung3D8Gb_extened.cfg Throws Floating point exception (core dumped)
sample_config_files/diff_ddr3_cache.cfg Quits with num_clk should be greater than zero!
sample_config_files/ddr3_cache.cfg Throws Segmentation fault (core dumped)
sample_config_files/lpddr3_cache.cfg Throws Segmentation fault (core dumped)
sample_config_files/wideio_cache.cfg Throws Segmentation fault (core dumped)

Backtrace of Segmentation fault:

Thread 1 "cacti" received signal SIGSEGV, Segmentation fault.
0x0000000000473d8e in find_all_bobs (memcad_params=0x0) at memcad.cc:396
396                     int last_bw =(*memcad_all_channels)[0]->bandwidth;
(gdb) bt
#0  0x0000000000473d8e in find_all_bobs (memcad_params=0x0) at memcad.cc:396
#1  0x0000000000474a1e in solve_memcad (memcad_params=0x7fffffff3310) at memcad.cc:595
#2  0x00000000004207c3 in cacti_interface (infile_name=...) at io.cc:1344
#3  0x0000000000411fd3 in main (argc=0, argv=0x656c69666e692d) at main.cc:78

Segmentation fault happens after logging the Low-swing wire stats. So until that point some time/area/power components are already reported. But the segmentation fault makes me suspicious about the correctness. Could you check that?

Backtrace of Floating point exception

Program received signal SIGFPE, Arithmetic exception.
0x000000000045fe31 in IOTechParam::IOTechParam (this=0x6ad070, g_ip=0x1400000000, io_type1=32767, num_mem_dq=-51952, mem_data_width=0, num_dq=7000176, connection=0, num_loads=1, freq=0) at extio_technology.cc:918
918                     (num_dq/mem_data_width)/(g_ip->num_clk/2); 
(gdb) bt
#0  0x000000000045fe31 in IOTechParam::IOTechParam (this=0x6ad070, g_ip=0x1400000000, io_type1=32767, num_mem_dq=-51952, mem_data_width=0, num_dq=7000176, connection=0, num_loads=1, freq=0) at extio_technology.cc:918
#1  0x00000000004206cf in cacti_interface (infile_name=...) at io.cc:1292
#2  0x0000000000411fd3 in main (argc=0, argv=0x656c69666e692d) at main.cc:78

Floating point exception happens directly after printing the configs, so no results are reported.

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions