
R_C1363912821
Members-
Content Count
23 -
Joined
-
Last visited
About R_C1363912821
-
Rank
Junior Member
-
Print from UVM Library
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
yes thats the only problem. I am running simulation on windows (Questa 10.0a), can this be a reason? -
Print from UVM Library
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
There is nothing like non UVM, the prints that are out of sync are from the UVM library # UVM_ERROR C:Users # RCDocumentsUVMuvmsrc/base/uvm_phases.svh(1860) @ 1000: reporter [PH_TIMEOUT] Phase timeout of 1000 hit, phase 'run' ready to end The issue is that it does not display the path properly (i.e. path is without '/') -
Print from UVM Library
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Hi Jadec, Thanks for providing the example, the issue is that i dont have a C compiler so i get error "# ** Fatal: (vsim-7019) Can't locate a C compiler for compilation of DPI export tasks/functions." is there any other solution? -
About the UVM objections.
R_C1363912821 replied to hugemx830202's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Hi Bart, Thanks for the reply, yes i was looking of UVM 1.0p1. BR, RC -
About the UVM objections.
R_C1363912821 replied to hugemx830202's topic in UVM (Pre-IEEE) Methodology and BCL Forum
The behavior when raise_objection is present (drop_objection removed) is expected, the point to confirm is when raise_objection is not used the sequence is not started and all subsequent sub-phase (reset, configure, main,....) and report phase are executed. Does this means that raise_objection is mandatory? -
About the UVM objections.
R_C1363912821 replied to hugemx830202's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Hi MEIXIAO, Your observation is correct, I had a working example which does not start sequence when raise_objection is removed, As per raise_objection description in UVM Reference ‘Raise an objection to ending this phase’, which seems to be appropriate here. When raise_objection is commented, this phase ends immediately and all other phases are executed after this (including report). Another case when only drop_objection is commented (with raise_objection in place), the sequence is executed (including body and post_body) and the run phase exists after default timeout with message # UVM_ERROR: /base/uvm_phases.svh(1855) @ 9200000000000: reporter [PH_TIMEOUT] Default phase timeout of 9200000000000 hit. All processes are waiting, indicating a probable testbench issue. Phase 'run' ready to end Hope this helps. UVM Developers, Please update if this behavior is intentional or not. Thanks, RC -
uvm_heartbeat
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Hi Kathleen, Thanks for providing example, it is really useful. I have a heartbeat example that was not working, hopefully I should be able to find out the root cause now. When I ran the example (provided by you) using Questa 10.0a there is a message # UVM_FATAL @ 1000: uvm_test_top [HBFAIL] Did not recieve an update of hb_obj on any component since last event trigger at time 500. The list of registered components is: # uvm_test_top.parent_0.child_0 # uvm_test_top.parent_0.child_1 # uvm_test_top.parent_0.child_2 at end which comes up to signal no activity, is there a way to have message to show that components are Active (instead of print for no activity). Thanks, RC -
Print from UVM Library
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Jadec, can you please provide example or steps required for this? -
Uvm_dump_cmdline_args
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Thanks Jadec, so the UVM reference has to be updated to make sure it is in sync with implementation. -
As per UVM Reference "+UVM_DUMP_CMDLINE_ARGS allows the user to dump all command line arguments to the reporting mechanism. The output in is tree format". When this is used with Questa 10.0a (on Windows) the output is not in tree format, only print that comes after adding this is # UVM_INFO @ 0: reporter [DUMPARGS] LoadDesign +UVM_TESTNAME=my_test +UVM_VERBOSITY=UVM_LOW +UVM_TIMEOUT=10000,NO +UVM_DUMP_CMDLINE_ARGS +UVM_OBJECTION_TRACE -do vsim.do -l rc.log -c -suppress 3829 -sv_lib ./uvm_dpi top There is nothing like tree format. Is this expected? if not the issue is with Simulator or UVM. Thanks, RC
-
Can someone provide example of heartbeat and how to check if components are alive?
-
Hi All, Internal prints from UVM library are not in Sync with other prints. e.g. the UVM_ERROR print in log below (For Questa 10.0a running on windows) is out of sync as it displays the complete path of file from which print comes (that too without '/'): # UVM_ERROR C:Users # RCDocumentsUVMuvmsrc/base/uvm_phases.svh(1860) @ 1000: reporter [PH_TIMEOUT] Phase timeout of 1000 hit, phase 'run' ready to end # UVM_WARNING @ 1000: run [OBJTN_CLEAR] Object 'common.run' cleared objection counts for run Is there to way to modify the display so that all prints look alike. Thanks, RC
-
Hi Bart, Setting verbosity to UVM_FULL also works well with Questa (Log as below), Thanks for suggestion. ------------------------------- # UVM_INFO C:Users/seq/uvm_sequencer_base.svh(1343) @ 0: uvm_test_top.m_env.m_sequencer [PHASESEQ] No default phase sequence for phase 'run' # UVM_INFO C:Users/seq/uvm_sequencer_base.svh(1343) @ 0: uvm_test_top.m_env.m_sequencer [PHASESEQ] No default phase sequence for phase 'pre_reset' ... # UVM_INFO C:Users/seq/uvm_sequencer_base.svh(1343) @ 0: uvm_test_top.m_env.m_sequencer [PHASESEQ] Starting default sequence 'my_sequence' for phase 'main' ------------------------------- Gordon, Default sequence is not displayed even when uvm_top.print() is called from end_of_elaboration_phase(), please check Janick’s reply at http://www.uvmworld.org/forums/showthread.php?168-enable_print_topology so the suggestion by Bart seems to be appropriate. Does anyone has any other idea? BR, RC
-
enable_print_topology
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Janick, Thanks for update. -
enable_print_topology
R_C1363912821 replied to R_C1363912821's topic in UVM (Pre-IEEE) Methodology and BCL Forum
Thanks for quick response Janick. It works, however there are few differences when UVM topology is compared with OVM topology e.g. for sequencer default sequence, count, sequences associated with sequencer are missing. OVM and UVM topology are copied below (for similar environment): # ---------------------------------------------------------------------- # Name Type Size Value # ---------------------------------------------------------------------- # ovm_test_top my_test - ovm_test_top@1 # m_env my_env - m_env@3 # m_driver my_driver - m_driver@5 # rsp_port ovm_analysis_port - rsp_port@9 # sqr_pull_port ovm_seq_item_pull_+ - sqr_pull_port@7 # m_sequencer my_sequencer - m_sequencer@11 # rsp_export ovm_analysis_export - rsp_export@13 # seq_item_export ovm_seq_item_pull_+ - seq_item_export@37 # default_sequence string 19 ovm_random_sequence # count integral 32 -1 # max_random_count integral 32 'd10 # sequences array 3 - # [0] string 19 ovm_random_sequence # [1] string 23 ovm_exhaustive_sequ+ # [2] string 19 ovm_simple_sequence # max_random_depth integral 32 'd4 # num_last_reqs integral 32 'd1 # num_last_rsps integral 32 'd1 # ---------------------------------------------------------------------- # ------------------------------------------------------------ # Name Type Size Value # ------------------------------------------------------------ # uvm_test_top my_test - @455 # m_env my_env - @464 # m_driver my_driver - @472 # rsp_port uvm_analysis_port - @487 # sqr_pull_port uvm_seq_item_pull_port - @479 # m_sequencer my_sequencer - @495 # rsp_export uvm_analysis_export - @502 # seq_item_export uvm_seq_item_pull_imp - @596 # arbitration_queue array 0 - # lock_queue array 0 - # num_last_reqs integral 32 'd1 # num_last_rsps integral 32 'd1 # ------------------------------------------------------------ Do i need to modify something to display that?