b498d99338
GNU ld has a counterintuitive lang_propagate_lma_regions rule.
```
// .foo's LMA region is propagated to .bar because their VMA region is the same,
// and .bar does not have an explicit output section address (addr_tree).
.foo : { *(.foo) } >RAM AT> FLASH
.bar : { *(.bar) } >RAM
// An explicit output section address disables propagation.
.foo : { *(.foo) } >RAM AT> FLASH
.bar . : { *(.bar) } >RAM
```
In both cases, lld thinks .foo's LMA region is propagated and
places .bar in the same PT_LOAD, so lld diverges from GNU ld w.r.t. the
second case (lma-align.test).
This patch changes Writer<ELFT>::createPhdrs to disable propagation
(start a new PT_LOAD). A user of the first case can make linker scripts
portable by explicitly specifying `AT>`. By contrast, there was no
workaround for the old behavior.
This change uncovers another LMA related bug in assignOffsets() where
`ctx->lmaOffset = 0;` was omitted. It caused a spurious "load address
range overlaps" error for at2.test
The new PT_LOAD rule is complex. For convenience, I listed the origins of some subexpressions:
* rL323449: `sec->memRegion == load->firstSec->memRegion`; linkerscript/at3.test
* D43284: `load->lastSec == Out::programHeaders` (don't start a new PT_LOAD after program headers); linkerscript/at4.test
* D58892: `sec != relroEnd` (start a new PT_LOAD after PT_GNU_RELRO)
Reviewed By: psmith
Differential Revision: https://reviews.llvm.org/D74297
45 lines
2.0 KiB
Plaintext
45 lines
2.0 KiB
Plaintext
# REQUIRES: x86
|
|
# RUN: llvm-mc -filetype=obj -triple=x86_64-unknown-linux %p/Inputs/at2.s -o %t.o
|
|
# RUN: ld.lld -o %t.exe %t.o --script %s
|
|
# RUN: llvm-readelf -l %t.exe | FileCheck %s
|
|
# RUN: llvm-objdump -section-headers %t.exe | FileCheck %s --check-prefix=SECTIONS
|
|
|
|
MEMORY {
|
|
AX (ax) : ORIGIN = 0x2000, LENGTH = 0x100
|
|
AW (aw) : ORIGIN = 0x3000, LENGTH = 0x100
|
|
FLASH (ax) : ORIGIN = 0x6000, LENGTH = 0x100
|
|
RAM (aw) : ORIGIN = 0x7000, LENGTH = 0x100
|
|
}
|
|
|
|
SECTIONS {
|
|
.foo1 : { *(.foo1) } > AX AT>FLASH
|
|
## In GNU ld, .foo1's LMA region is propagated to .foo2 because their VMA region
|
|
## is the same and .foo2 does not set an explicit address.
|
|
## lld sets .foo2's LMA region to null.
|
|
.foo2 : { *(.foo2) } > AX
|
|
|
|
.bar1 : { *(.bar1) } > AW
|
|
.bar2 : { *(.bar2) } > AW AT > RAM
|
|
.bar3 . : { *(.bar3) } > AW
|
|
.bar4 : { *(.bar4) } > AW AT >RAM
|
|
}
|
|
|
|
# CHECK: Type Offset VirtAddr PhysAddr FileSiz MemSiz Flg Align
|
|
# CHECK-NEXT: LOAD 0x001000 0x0000000000002000 0x0000000000006000 0x000008 0x000008 R E 0x1000
|
|
# CHECK-NEXT: LOAD 0x001008 0x0000000000002008 0x0000000000002008 0x000008 0x000008 R E 0x1000
|
|
# CHECK-NEXT: LOAD 0x002000 0x0000000000003000 0x0000000000003000 0x000008 0x000008 RW 0x1000
|
|
# CHECK-NEXT: LOAD 0x002008 0x0000000000003008 0x0000000000007000 0x000008 0x000008 RW 0x1000
|
|
# CHECK-NEXT: LOAD 0x002010 0x0000000000003010 0x0000000000003010 0x000008 0x000008 RW 0x1000
|
|
# CHECK-NEXT: LOAD 0x002018 0x0000000000003018 0x0000000000007008 0x000008 0x000008 RW 0x1000
|
|
|
|
# SECTIONS: Sections:
|
|
# SECTIONS-NEXT: Idx Name Size VMA
|
|
# SECTIONS-NEXT: 0 00000000 0000000000000000
|
|
# SECTIONS-NEXT: 1 .foo1 00000008 0000000000002000
|
|
# SECTIONS-NEXT: 2 .foo2 00000008 0000000000002008
|
|
# SECTIONS-NEXT: 3 .text 00000000 0000000000002010
|
|
# SECTIONS-NEXT: 4 .bar1 00000008 0000000000003000
|
|
# SECTIONS-NEXT: 5 .bar2 00000008 0000000000003008
|
|
# SECTIONS-NEXT: 6 .bar3 00000008 0000000000003010
|
|
# SECTIONS-NEXT: 7 .bar4 00000008 0000000000003018
|