David Fang

Results 33 issues of David Fang

Removing a symbolic link looks like removing the link's destination (but perhaps should not?). Demonstration: ``` $ touch foo $ ln -s foo bar $ fsatrace erwdtmq /dev/stdout -- rm...

https://github.com/fangism/hackt/blob/e77f05bd2382f40c9493b86b9b32ef9a009b3263/src/Object/inst/instance_alias.tcc#L777 has appeared in a stacktrace in a large design that stack-overflowed during the load of an elaborated .haco-c file (happened to be in cosim). Rewriting that function should greatly...

Undefined symbols: "_lt_dladdsearchdir", referenced from: HAC::chpsim::parse_command_options(int, char*_, HAC::chpsim_options&)in libhacktexec.a(chpsim.o) ld: symbol(s) not found collect2: ld returned 1 exit status make[2]: *_\* [hackt] Error 1 This was expected to have been...

build-portability
workaround

Would like to implement an exhaustive prover for stability/non-interference using prsim-based state exploration and coverage, and reachability analysis. The legacy tool that inspires this is "prlint." Without this, we rely...

feature

During build on x86_64-linux using ``` Linux 3.2.0-23-generic #36-Ubuntu SMP gcc version 4.9.2 (Ubuntu 4.9.2-0ubuntu1~12.04) GNU ld (GNU Binutils for Ubuntu) 2.22 ``` I see linker errors like: ``` /bin/sh...

build-portability

e.g. c1of2 forward, c1of2 backward, where acknowledge carries data, not just a single wire. Workaround for lack of this feature: channel adapters on the forward and reverse halves.

simulator
feature

Currently, instances and parameter collections are internally limited to 4-dimensional arrays. There's no good reason for this anymore, when we have the data structures to support arbitrary N-dimensions. We can...

language front-end

Consider the following Verilog (used for cosimulation) that takes a file name from a plusarg: ``` initial begin reg [128:0] prs_obj; if (!$value$plusargs("PRSIM_OBJ=%s", prs_obj)) begin prs_obj = "Missing required +PRSIM_OBJ=...

simulator
workaround

Could use documentation for spice\* helper scripts. Should they go into hacknet documentation, even though they only operate on generated output?

documentation

The texinfo documentation for hacknet is already organized in sections, but the command-line -H option is alphabetically sorted across all options. Would be nice if this were also arranged in...

documentation