Re: [PATCH] doc: Add stapsdt provider documentation

On Tue, Nov 18, 2025 at 07:15:16PM -0500, Eugene Loh wrote:
> On 11/18/25 07:43, Alan Maguire wrote:
> 
> > On 18/11/2025 04:30, [email protected] wrote:
> > > From: Eugene Loh <[email protected]>
> > > 
> > > Signed-off-by: Eugene Loh <[email protected]>
> > looks great, a few small suggestions below, but
> > 
> > Reviewed-by: Alan Maguire <[email protected]>
> 
> Thanks, but...
> 
> > > diff --git a/doc/userguide/reference/dtrace_providers_stapsdt.md b/doc/userguide/reference/dtrace_providers_stapsdt.md
> > > +    STAP_PROBE3(myprovider, myprobe, argc, argv[0], 18);
> > > +    if (myprovider_myprobe_semaphore)
> > > +        printf("the probe is enabled\n");
> > > +    else
> > > +        printf("the probe is not enabled\n");
> > I found the example a bit confusing. I would have thought:
> 
> (fwiw.  I think it's cool you can check it anyhow, but yeah your example
> makes more sense.)
> 
> > > +Notice that the instrumented source code includes the header file
> > > +`<sys/sdt.h>`.  On Oracle Linux, this file is installed
> > > +by the RPM package `systemtap-sdt-devel`, but the package also
> > > +installs `/usr/bin/dtrace`.  So be sure to have `/usr/sbin`
> > > +in front of `/usr/bin` in your path, or explicitly specify
> > > +`/usr/sbin/dtrace` whenever you want to use `dtrace`.
> > > +
> > Would be good to mention that a double underscore "__" gets converted to
> > a dash (-) in provider/probe names too.
> 
> ... but what does this have to do with stapsdt?  I thought that was a
> general DTrace thing that would not need to be called out specially here.

I agree - this is standard in DTrace so I don't think it needs to be pointed
out.  Is it done that way in systemtap as well?  If so, no need to mention it.
If systemtap does *not* do that, I would mention it here because that would be
a deviation from systemtap.