Hi Tom,
On Fri, Oct 23, 2020 at 09:58:34PM -0600, Tom Tromey wrote:
> PR 26773 points out that some sleb128 values are decoded incorrectly.
> Looking into this, I found some other unusual cases as well.
>
> In this patch, I chose to try to handle weird leb128 encodings by
> preserving their value
Mark> Clearly DW_FORM_strx3 and DW_FORM_addrx3 aren't used much
Mark> (given that it is an index value between larger than 65536).
I don't even really see the need for a 3-byte form.
Anyway if it helps, gdb gets this wrong too -- gdb always uses
little-endian, which seems wrong, but I didn't know
On Fri, Oct 23, 2020 at 04:51:10PM -0600, Tom Tromey wrote:
> The read_3ubyte_unaligned_inc macro calls read_2ubyte_unaligned, but
> it should call read_3ubyte_unaligned.
Eep. And read_3ubyte_unaligned is actually is tricky to get right if
you want to correctly read 3-byte unsigned integers given
