Nintendo Switch Brew

Changes

NSO (view source)

Revision as of 18:04, 22 February 2020

734 bytes added ,  18:04, 22 February 2020
no edit summary
Line 1: Line 1: −
[[Category:File formats]]
   
NSO is the main executable format.
 
NSO is the main executable format.
    
It starts with the "NSO" header and mainly describes .text, .rodata, and .data segments (like a short-form of ELF program headers). If the segments are compressed, they are compressed using LZ4.
 
It starts with the "NSO" header and mainly describes .text, .rodata, and .data segments (like a short-form of ELF program headers). If the segments are compressed, they are compressed using LZ4.
   −
= NSO Header =
+
= Header =
 
{| class="wikitable" border="1"
 
{| class="wikitable" border="1"
 
|-
 
|-
Line 17: Line 16:  
| 0x4
 
| 0x4
 
| 4
 
| 4
−
| NSO Version (Always 0)
+
| Version (always 0)
 
|-
 
|-
 
| 0x8
 
| 0x8
 
| 4
 
| 4
−
| Reserved (Unused)
+
| Reserved
 
|-
 
|-
 
| 0xC
 
| 0xC
 
| 4
 
| 4
−
| Flags, bit 0-2: (.text, .rodata and .data) section is compressed, bit 3-5: check section hash when loading
+
| [[#Flags|Flags]]
 
|-
 
|-
 
| 0x10
 
| 0x10
 
| 0xC
 
| 0xC
−
| .text SegmentHeader
+
| [[#SegmentHeader|SegmentHeader]] for .text
 
|-
 
|-
 
| 0x1C
 
| 0x1C
 
| 0x4  
 
| 0x4  
−
| Module offset (calculated by sizeof(header))
+
| ModuleNameOffset (calculated by sizeof(header))
 
|-
 
|-
 
| 0x20
 
| 0x20
 
| 0xC
 
| 0xC
−
| .rodata SegmentHeader
+
| [[#SegmentHeader|SegmentHeader]] for .rodata
 
|-
 
|-
 
| 0x2C
 
| 0x2C
 
| 0x4
 
| 0x4
−
| Module file size
+
| ModuleNameSize
 
|-
 
|-
 
| 0x30
 
| 0x30
 
| 0xC
 
| 0xC
−
| .data SegmentHeader
+
| [[#SegmentHeader|SegmentHeader]] for .data
 
|-
 
|-
 
| 0x3C
 
| 0x3C
 
| 0x4
 
| 0x4
−
| bssSize
+
| BssSize
 
|-
 
|-
 
| 0x40
 
| 0x40
 
| 0x20
 
| 0x20
−
| Value of "build id" from ELF's GNU .note section. Contains variable sized digest, up to 32bytes.
+
| [[#ModuleId|ModuleId]]
 
|-
 
|-
 
| 0x60
 
| 0x60
 
| 0x4
 
| 0x4
−
| .text compressed size
+
| TextFileSize (.text compressed size)
 
|-
 
|-
 
| 0x64
 
| 0x64
 
| 0x4
 
| 0x4
−
| .rodata compressed size
+
| RoFileSize (.rodata compressed size)
 
|-
 
|-
 
| 0x68
 
| 0x68
 
| 0x4
 
| 0x4
−
| .data compressed size
+
| DataFileSize (.data compressed size)
 
|-
 
|-
 
| 0x6C
 
| 0x6C
 
| 0x1C
 
| 0x1C
−
| Reserved (Padding)
+
| Reserved
 
|-
 
|-
 
| 0x88
 
| 0x88
 
| 0x8
 
| 0x8
−
| .rodata-relative extents of .api_info
+
| [[#SegmentHeaderRelative|SegmentHeaderRelative]] for .api_info
 
|-
 
|-
 
| 0x90
 
| 0x90
 
| 0x8
 
| 0x8
−
| .rodata-relative extents of .dynstr
+
| [[#SegmentHeaderRelative|SegmentHeaderRelative]] for .dynstr
 
|-
 
|-
 
| 0x98
 
| 0x98
 
| 0x8
 
| 0x8
−
| .rodata-relative extents of .dynsym
+
| [[#SegmentHeaderRelative|SegmentHeaderRelative]] for .dynsym
 
|-
 
|-
 
| 0xA0
 
| 0xA0
−
| 0x20 * 3
+
| 0x20
−
| SHA256 hashes over the decompressed sections using the above byte-sizes: .text, .rodata, and .data.
+
| TextHash (SHA256 hash over the decompressed .text section using the above size)
 +
|-
 +
| 0xC0
 +
| 0x20
 +
| RoHash (SHA256 hash over the decompressed .rodata section using the above size)
 +
|-
 +
| 0xE0
 +
| 0x20
 +
| DataHash (SHA256 hash over the decompressed .data section using the above size)
 
|-
 
|-
 
| 0x100
 
| 0x100
−
|
+
| Variable
 
| Compressed sections
 
| Compressed sections
 
|}
 
|}
Line 94: Line 101:  
Most data in Switch binaries are standard ELF structures, however some are custom.
 
Most data in Switch binaries are standard ELF structures, however some are custom.
 
For example, the MOD header is essentially a replacement for a PT_DYNAMIC program header.
 
For example, the MOD header is essentially a replacement for a PT_DYNAMIC program header.
 +
 +
== Flags ==
 +
{| class="wikitable" border="1"
 +
|-
 +
! Bits
 +
! Description
 +
|-
 +
| 0
 +
| TextCompress (.text section is compressed)
 +
|-
 +
| 1
 +
| RoCompress (.rodata section is compressed)
 +
|-
 +
| 2
 +
| DataCompress (.data section is compressed)
 +
|-
 +
| 3
 +
| TextHash (.text hash must be checked when loading)
 +
|-
 +
| 4
 +
| RoHash (.rodata hash must be checked when loading)
 +
|-
 +
| 5
 +
| DataHash (.data hash must be checked when loading)
 +
|}
    
== SegmentHeader ==
 
== SegmentHeader ==
Line 112: Line 144:  
| 0x8
 
| 0x8
 
| 0x4
 
| 0x4
−
| DecompressedSize
+
| Size (decompressed)
 
|}
 
|}
   −
== .rodata-relative extent ==
+
== ModuleId ==
 +
Value of "build id" from ELF's GNU .note section. Contains variable sized digest, up to 32bytes.
 +
 
 +
== SegmentHeaderRelative ==
 
{| class="wikitable" border="1"
 
{| class="wikitable" border="1"
 
|-
 
|-
Line 124: Line 159:  
| 0x0
 
| 0x0
 
| 4
 
| 4
−
| RegionRoDataOffset
+
| Offset
 
|-
 
|-
 
| 0x4
 
| 0x4
 
| 4
 
| 4
−
| RegionSize
+
| Size
 
|}
 
|}
 +
 +
Offset is relative to the .rodata section.
    
== MOD ==
 
== MOD ==
−
All offsets are signed 32bit values relative to the magic field.
  −
The 32bits at image base + 4 must point to the magic field.
  −
The MOD structure is designed such that it can be placed at image base and point to itself.
  −
The 2 fields preceding the magic field get copied around with the structure, even if it is relocated to somewhere besides the image base. If MOD is not located at image base, the value at offset 4 must still point to the MOD magic. In the case of .text being at image base, this implies that the first instruction can only be an unconditional branch over the offset literal.
   
{| class="wikitable" border="1"
 
{| class="wikitable" border="1"
 
|-
 
|-
Line 144: Line 177:  
| 0x00
 
| 0x00
 
| 4
 
| 4
−
| ZeroPadding
+
| Reserved
 
|-
 
|-
 
| 0x04
 
| 0x04
 
| 4
 
| 4
−
| MagicOffset. Always 8 (so it works when MOD is at image_base + 0).
+
| MagicOffset (always 8, so it works when MOD is at image_base + 0)
 
|-
 
|-
 
| 0x08
 
| 0x08
Line 176: Line 209:  
| 0x20
 
| 0x20
 
| 4
 
| 4
−
| offset to runtime-generated module object. typically equal to .bss base.
+
| Offset to runtime-generated module object (typically equal to .bss base)
 
|}
 
|}
   −
=Arguments=
+
All offsets are signed 32bit values relative to the magic field.
−
Loader maps memory and writes the [[Loader_services#AddProcessToLaunchQueue|arguments]] to {end of rwdata section specified by last SegmentHeader}. Official processes use argdata_addr = {page-aligned _end}. svcQueryMemory is used by official sw to verify that argdata_addr is mapped RW, since this memory is only mapped when arguments are specified via that command. Afterwards, official sw aligns the argdata_addr to 4-bytes. Structure located at argdata_addr:
+
The 32bits at image base + 4 must point to the magic field.
 +
The MOD structure is designed such that it can be placed at image base and point to itself.
 +
The 2 fields preceding the magic field get copied around with the structure, even if it is relocated to somewhere besides the image base. If MOD is not located at image base, the value at offset 4 must still point to the MOD magic. In the case of .text being at image base, this implies that the first instruction can only be an unconditional branch over the offset literal.
 +
 
 +
= Arguments =
 +
[[Loader_services|Loader]] maps memory and writes the [[Loader_services#SetProgramArgument|arguments]] to {end of rwdata section specified by last SegmentHeader}. Official processes use argdata_addr = {page-aligned _end}. svcQueryMemory is used by official sw to verify that argdata_addr is mapped RW, since this memory is only mapped when arguments are specified via that command. Afterwards, official sw aligns the argdata_addr to 4-bytes.
    +
The structure located at argdata_addr is as follows:
 
{| class="wikitable" border="1"
 
{| class="wikitable" border="1"
 
|-
 
|-
Retrieved from "https://switchbrew.org/wiki/Special:MobileDiff/8770"