[libre-riscv-dev] daily kan-ban update 26jun2020
mtnolan2640 at gmail.com
Fri Jun 26 20:38:55 BST 2020
On 6/26/20 3:29 PM, Luke Kenneth Casson Leighton wrote:
> On Fri, Jun 26, 2020 at 8:17 PM Michael Nolan <mtnolan2640 at gmail.com> wrote:
>> On 6/26/20 3:15 PM, Luke Kenneth Casson Leighton wrote:
>>> tck-tck.... i set it to 1 when pi.is_ld_i is set, still getting
>>> spurious responses.
>>> l'll open up the graphviz and the vcd file, see what's going on.
>> I tried setting it when there was a load previously and x_busy_o is
>> clear. It works for one load, but not two back to back.
> ok that will be becauuuse.... x_busy_o is not being passed out to pi.busy_o.
Nah, lsmem is never busy, it responds to a read on the following cycle.
I think the test case starts the second read on the same cycle as the
first read finishes (which is fine), but it sees that ld.ok is true and
assumes that it means the second read is finished. I think I got it
More information about the libre-riscv-dev