Commit 27b81b46 authored by Simon Marlow's avatar Simon Marlow

improve comments about x86-64 relative-offset hackery

parent 4414ec29
......@@ -760,7 +760,8 @@ pprDataItem lit
-- Fortunately we're assuming the small memory model, in which
-- all such offsets will fit into 32 bits, so we have to stick
-- to 32-bit offset fields and modify the RTS appropriately
-- (see InfoTables.h).
-- See Note [x86-64-relative] in includes/InfoTables.h
ppr_item I64 x
| isRelativeReloc x =
......@@ -236,6 +236,8 @@ sub init_TARGET_STUFF {
# where x is in the text section and y in the rodata section.
# It works if y is in the text section, though. This is probably
# going to cause difficulties for PIC, I imagine.
# See Note [x86-64-relative] in includes/InfoTables.h
$T_HDR_relrodata= "\.text\n\t\.align 8\n";
$T_HDR_closure = "\.data\n\t\.align 8\n";
......@@ -216,11 +216,21 @@ typedef struct StgLargeSRT_ {
relative to the info pointer, so that we can generate
position-independent code.
Note [x86-64-relative]
There is a complication on the x86_64 platform, where pointeres are
64 bits, but the tools don't support 64-bit relative relocations.
However, the default memory model (small) ensures that all symbols
have values in the lower 2Gb of the address space, so offsets all
fit in 32 bits. Hence we can use 32-bit offset fields.
When going via-C, the mangler arranges that we only generate
relative relocations between symbols in the same segment (the text
segment). The NCG, however, puts things in the right sections and
uses 32-bit relative offsets instead.
Somewhere between binutils-2.16.1 and binutils-,
support for 64-bit PC-relative relocations was added, so maybe this
hackery can go away sometime.
------------------------------------------------------------------------- */
#if x86_64_TARGET_ARCH
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment