Re: [RFA] Introduce notion of "search name"

[Daniel Jacobowitz <drow@false.org>]

On Thu, May 13, 2004 at 05:30:23AM -0400, Paul Hilfinger wrote:
> 
> Oh great.  I walk into my office to find skull fragments all over the
> floor, but still no final decision.  Ah well.  Risking exposure to
> stray prions, I submit the latest version of this patch below.
> 
> I am actually sympathetic to Andrew's ideas here, since with some
> version of the extra abstraction he suggests, I could avoid all
> permanent storage for demangled names.  However, it WOULD be a
> substantial change and speaking for ACT, we'd really, really first
> like to get some substantial portion of the Ada changes integrated
> into (and active) in the public GDB version.

This patch is OK with me; let's see if Jim likes it.  Thanks for
following through; you've been a real sport about this.

> 
> Paul Hilfinger
> 
> 2004-05-13  Paul N. Hilfinger  <hilfinger@gnat.com>
> 
> 	* symtab.h (SYMBOL_SEARCH_NAME): New definition.
> 	(SYMBOL_MATCHES_SEARCH_NAME): New definition.
> 	(symbol_search_name): Declare.
> 	
> 	* dictionary.c (iter_name_first_hashed): Match on SYMBOL_SEARCH_NAME.
> 	(iter_name_next_hashed): Ditto.
> 	(iter_name_next_linear): Ditto.
> 	(insert_symbol_hashed): Hash on SYMBOL_SEARCH_NAME.
> 
> 	* symtab.c (lookup_partial_symbol): Assume symbols ordered by 
> 	search name, using SYMBOL_SEARCH_NAME and SYMBOL_MATCHES_SEARCH_NAME.
> 	(symbol_search_name): New function.
> 	
> 	* symfile.c (compare_psymbols): Order by SYMBOL_SEARCH_NAME.
> 	
> 	* minsyms.c (build_minimal_symbol_hash_tables): Change
> 	test for adding to demangled hash table to check for difference
> 	between SYMBOL_SEARCH_NAME and SYMBOL_LINKAGE_NAME.

-- 
Daniel Jacobowitz