Internals of -readnow implementation

Internals of -readnow implementation

[Jan Vrany via Gdb]

Hi, 

looking at how -readnow is implemented, I'm puzzled
why implementation of readnow_functions inherits/user
dwarf2_base_index_functions?

My understanding is that if -readnow is specified, then
all compunit symbtabs in that objfile are fully read.
It seems to me that in theory one can implement it "directly" 
by traversing compunit symtabs, its blockvector and symbols 
(rather than using dwarf2_per_objfile and alike). 

Is that because it was just easier to implement it this way
or is there a reason why it has to go through DWARF reader
structures?

Thanks, 
Jan

Re: Internals of -readnow implementation

[Tom Tromey]

Jan> looking at how -readnow is implemented, I'm puzzled
Jan> why implementation of readnow_functions inherits/user
Jan> dwarf2_base_index_functions?

It's mostly a convenience, because that class implements some methods.

Jan> My understanding is that if -readnow is specified, then
Jan> all compunit symbtabs in that objfile are fully read.
Jan> It seems to me that in theory one can implement it "directly" 
Jan> by traversing compunit symtabs, its blockvector and symbols 
Jan> (rather than using dwarf2_per_objfile and alike). 

Jan> Is that because it was just easier to implement it this way
Jan> or is there a reason why it has to go through DWARF reader
Jan> structures?

There's no deep reason.

For this bug:

    https://sourceware.org/bugzilla/show_bug.cgi?id=33554

... I am planning to reimplement it as you say: just search things
directly without regard to the DWARF data structures.

-readnow is weird because it's mostly there as a debugging aid, and
sometimes a workaround for other bugs.  While I have recommended it to
users, I don't really like doing so.  Occasionally I think about
removing it.

Tom

Re: Internals of -readnow implementation

[Jan Vrany via Gdb]

On Tue, 2025-11-11 at 14:18 -0700, Tom Tromey wrote:
> Jan> looking at how -readnow is implemented, I'm puzzled
> Jan> why implementation of readnow_functions inherits/user
> Jan> dwarf2_base_index_functions?
> 
> It's mostly a convenience, because that class implements some methods.

I see, thanks. 
> 
> Jan> My understanding is that if -readnow is specified, then
> Jan> all compunit symbtabs in that objfile are fully read.
> Jan> It seems to me that in theory one can implement it "directly" 
> Jan> by traversing compunit symtabs, its blockvector and symbols 
> Jan> (rather than using dwarf2_per_objfile and alike). 
> 
> Jan> Is that because it was just easier to implement it this way
> Jan> or is there a reason why it has to go through DWARF reader
> Jan> structures?
> 
> There's no deep reason.
> 
> For this bug:
> 
>     https://sourceware.org/bugzilla/show_bug.cgi?id=33554
> 
> ... I am planning to reimplement it as you say: just search things
> directly without regard to the DWARF data structures.

Yes! I encountered the same issue when working on Python JIT API
and came to the same realization.

In fact, I already started working on it and got relatively far
in replacing current implementation of readnow_functions but still 
have loose ends especially in ::search method. It seems to me that the 
implementations not always match its comment.

Jan

> 
> -readnow is weird because it's mostly there as a debugging aid, and
> sometimes a workaround for other bugs.  While I have recommended it to
> users, I don't really like doing so.  Occasionally I think about
> removing it.
> 
> Tom