Re: [PATCH 2/6] Set section indices when symbols are made

[Tom Tromey <tromey@adacore.com>]

>>>>> "Simon" == Simon Marchi <simark@simark.ca> writes:

Simon> On 1/18/23 10:30, Tom Tromey via Gdb-patches wrote:
>> Most places in gdb that create a new symbol will apply a section
>> offset to the address.  It seems to me that the choice of offset here
>> is also an implicit choice of the section.  This is particularly true
>> if you examine fixup_section, which notes that it must be called
>> before such offsets are applied -- meaning that if any such call has
>> an effect, it's purely by accident.
>> 
>> This patch cleans up this area by tracking the section index and
>> applying it to a symbol when the address is set.  This is done for
>> nearly every case -- the remaining cases will be handled in later
>> patches.

Simon> So, if I rephrase it to make sure I understand: it's not logical to
Simon> apply the relocation to the symbol's address, without also setting the
Simon> symbol's section index, because the symbol's relocation value comes from
Simon> the section.  Does that sound right?

Yes, that's right.

Hypothetically, if you had a runtime linker that applied different
offsets to different sections (like, text gets +0x1000, data gets
+0x2000, bss gets +0x3000), then the code that applies these offsets
could sometimes be wrong.  But, since it's been this way a long time
and, AFAIK, there aren't problems with it, then IMO it follows that
setting the section from these choices is fine.

One other oddity I found here is that I don't think any code uses
anything other than SECT_OFF_TEXT for functions.  So maybe the code to
deal with the section index in a compunit symtab is not actually needed.

Tom