Re: [rtems commit] pc386: Add Edison base support

2015-01-07 Thread Joel Sherrill
On January 7, 2015 3:04:42 PM CST, Chris Johns wrote: >On 5/01/2015 6:45 am, Joel Sherril wrote: > >Why code like > >> +#if (BSP_IS_EDISON == 0) > >and ... > >> +#if (BSP_IS_EDISON == 1) > >this rather than > >#if !BSP_IS_EDISON > >and > >#if BSP_IS_EDISON > >The advantages are unclear

Re: [rtems commit] pc386: Add Edison base support

2015-01-07 Thread Chris Johns
On 5/01/2015 6:45 am, Joel Sherril wrote: Why code like +#if (BSP_IS_EDISON == 0) and ... +#if (BSP_IS_EDISON == 1) this rather than #if !BSP_IS_EDISON and #if BSP_IS_EDISON The advantages are unclear to me. Chris ___ devel mailin

RE: Trace interrupts with the capture engine

2015-01-07 Thread Jennifer Averett
A couple of points: 1. What we had discussed in the past was using rtems-tld to generate wrappers 2. This only works for simple vectored and doesn't account for PIC interrupts Jennifer Averett On-Line Applications Research From: Daniel Ceder

Trace interrupts with the capture engine

2015-01-07 Thread Daniel Cederman
Hello, I am looking into adding support to the capture engine to trace interrupts. One possible way, which seems generic and would not require any changes to the BSPs, is to replace the ISR for a specific interrupt with a wrapper that adds a record to the capture log before and after it calls