| 1 |
/* PR 25662: objcopy sets invalid sh_offset for the first section in a |
| 2 |
no_contents segment containing program headers. |
| 3 |
|
| 4 |
Several conditions are required for the bug to manifest: |
| 5 |
- The first loadable segment (which contains the program headers) must only |
| 6 |
contain SHT_NOBITS sections. .bss is the SHT_NOBITS section in this test. |
| 7 |
- The next loadable segment must have a !SHT_NOBITS loadable section. .data |
| 8 |
is the !SHT_NOBITS section in this test. |
| 9 |
- .bss must be positioned after .data in the executable file itself. |
| 10 |
- The size of .data must be such that the calculated VMA of the .bss |
| 11 |
section that follows it is not congruent with the file offset of .bss, |
| 12 |
modulo the p_align of its segment, i.e.: |
| 13 |
(VMA(.data) + sizeof(.data)) % (.bss_segment.p_align) != 0 |
| 14 |
This will force the sh_offset of .bss to be aligned so it appears within |
| 15 |
.data. |
| 16 |
- The size of .data must be larger than the program headers in the first |
| 17 |
loadable segment, so that the file offset of .bss is immediately |
| 18 |
after .data, and not padded to a valid alignment by the program headers. |
| 19 |
|
| 20 |
The bug originally only manifested for ELF targets, but there's no reason not |
| 21 |
to run this testcase for other file formats. */ |
| 22 |
|
| 23 |
.section .bss |
| 24 |
a: |
| 25 |
.zero 0x2 |
| 26 |
|
| 27 |
.section .data |
| 28 |
c: |
| 29 |
.zero 0x201 |
| 30 |
|
| 31 |
.section .text |
| 32 |
.global _start |
| 33 |
_start: |
| 34 |
.long 0 |