If a .s file is created which uses .long, .short or .byte directives to specify
the object code, instead of instruction mnenomics, the assembler gets the
address in the first entry in the table incorrect by several bytes.
E.g. using a test source
void foo()
{
}
and compiling this with gcc (-S op
--- Additional Comments From andyhhp at hotmail dot com 2010-07-09 12:44
---
Created an attachment (id=4877)
--> (http://sourceware.org/bugzilla/attachment.cgi?id=4877&action=view)
Test case demonstrating bug.
Included is:
main.c - fragment of code which causes the bug to occur
asm.s -
I am building an implementation of MIPS on an FPGA and am using GCC 4.5.0 with
binutils 2.20 to generate code for it. The processor does not exhibit branch or
load delay slots.
For debugging purposes, I am using the 'break' opcode but gcc does not know
this. As a result, i am using inline assemb
--- Additional Comments From ktietz at onevision dot com 2010-07-09 11:56
---
The failure isn't reasoned by shift reduces. The underlying problem you see
here is, that for IDs the keywords aren't recognized. So symbols named 'data',
or 'BASE', etc aren't recognized by grammer.
This beh
--- Additional Comments From daniel dot calcoen at cern dot ch 2010-07-09
10:14 ---
fixed by DJ Delorie on 3rd Jul 2010
--
What|Removed |Added
Status|NEW
