rust/zynq7000-rs
7
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

init commit

Browse Source
This commit is contained in:
Robin Müller
2025-02-19 11:00:04 +01:00
commit f842673e3a
104 changed files with 21595 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
zynq7000-rt/Cargo.lock generated Normal file
View File

@ -0,0 +1,7 @@
# This file is automatically @generated by Cargo.
# It is not intended for manual editing.
version = 4
[[package]]
name = "zynq-rt"
version = "0.1.0"

26
zynq7000-rt/Cargo.toml Normal file
View File

@ -0,0 +1,26 @@
[package]
name = "zynq7000-rt"
version = "0.1.0"
authors = ["Robin Mueller <muellerr@irs.uni-stuttgart.de>"]
edition = "2024"
description = "Run-time support for the Zynq7000 family of SoCs for running bare-metal applications"
homepage = "https://egit.irs.uni-stuttgart.de/rust/zynq7000-rs"
repository = "https://egit.irs.uni-stuttgart.de/rust/zynq7000-rs"
license = "MIT OR Apache-2.0"
keywords = ["no-std", "arm", "cortex-a", "amd", "zynq7000"]
categories = ["embedded", "no-std", "hardware-support"]
[dependencies]
cortex-a-rt = { git = "https://github.com/rust-embedded/cortex-ar", branch = "main", optional = true, features = ["vfp-dp"] }
cortex-ar = { git = "https://github.com/rust-embedded/cortex-ar", branch = "main", features = ["critical-section-single-core"] }
[features]
default = ["rt"]
tools = []
rt = ["dep:cortex-a-rt"]
[[bin]]
name = "table-gen"
path = "src/bin/table-gen.rs"
# Prevents default build
required-features = ["tools"]

27
zynq7000-rt/README.md Normal file
View File

@ -0,0 +1,27 @@
Zynq7000 Rust Run-Time Support
========
Startup code and minimal runtime for the AMD Zynq7000 SoC to write bare metal Rust code.
This run-time crate is strongly based on the
[startup code provided by AMD](https://github.com/Xilinx/embeddedsw/blob/master/lib/bsp/standalone/src/arm/cortexa9/gcc/boot.S).
One major difference is that the MMU table is specified as Rust code. There are also modification
to the stack setup code, because a different linker script is used.
This crate pulls in the [cortex-a-rt](https://github.com/us-irs/cortex-ar/tree/cortex-a-addition/cortex-a-rt)
crate to provide ARM vectors and the linker script.
## Features
- `rt` is a default feature which activates the run-time.
## Re-Generating the MMU table
The MMU table is a static flat map of 4096 entries for each 1 MB in the memory map.
It was generated using the `table-gen` binary tool.
You can re-run the tool using
```sh
cargo +stable --target <hostTarget> run --bin table-gen --no-default-features --features tools
```

2
zynq7000-rt/regen-table.sh Executable file
View File

@ -0,0 +1,2 @@
#!/bin/bash
cargo +stable run --target $(rustc -vV | grep host | cut -d ' ' -f2) --bin table-gen --no-default-features --features tools

285
zynq7000-rt/src/bin/table-gen.rs Normal file
View File

@ -0,0 +1,285 @@
use std::fs::File;
use std::io::Write;
use std::process::Command;
use zynq7000_rt::mmu::ONE_MB;
pub use zynq7000_rt::mmu::segments::*;
fn main() {
let file_path = "src/mmu_table.rs";
let file = File::create(file_path).expect("Failed to create file");
let mut offset = 0;
let attr_ddr = stringify!(section_attrs::DDR);
let attr_unassigned = stringify!(section_attrs::UNASSIGNED_RESERVED);
let attr_fpga_slaves = stringify!(section_attrs::FPGA_SLAVES);
let attr_shared_dev = stringify!(section_attrs::SHAREABLE_DEVICE);
let attr_sram = stringify!(section_attrs::SRAM);
let attr_qspi = stringify!(section_attrs::QSPI_XIP);
let attr_ocm_high = stringify!(section_attrs::OCM_MAPPED_HIGH);
assert_eq!(
1 + DDR_FULL_ACCESSIBLE
+ FPGA_SLAVE
+ FPGA_SLAVE
+ UNASSIGNED_0
+ IO_PERIPHS
+ UNASSIGNED_1
+ NAND
+ NOR
+ SRAM
+ SEGMENTS_UNASSIGNED_2
+ AMBA_APB
+ UNASSIGNED_3
+ QSPI_XIP
+ UNASSIGNED_4
+ OCM_MAPPED_HIGH,
4096
);
let mut buf_writer = std::io::BufWriter::new(file);
writeln!(
buf_writer,
"//! This file was auto-generated by table-gen.rs"
)
.unwrap();
writeln!(buf_writer, "use cortex_ar::mmu::L1Section;").unwrap();
writeln!(buf_writer, "use crate::mmu::{{section_attrs, L1Table}};").unwrap();
writeln!(buf_writer, "").unwrap();
writeln!(buf_writer, "/// MMU Level 1 Page table.").unwrap();
writeln!(buf_writer, "///").unwrap();
writeln!(
buf_writer,
"/// 4096 entries, each covering 1MB of the address space."
)
.unwrap();
writeln!(
buf_writer,
"pub const MMU_L1_PAGE_TABLE: L1Table = L1Table(["
)
.unwrap();
writeln!(
buf_writer,
"// First DDR segment, OCM memory (0x0000_0000 - 0x0010_0000)"
)
.unwrap();
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_ddr
)
.unwrap();
offset += ONE_MB;
writeln!(buf_writer, "// DDR memory (0x00100000 - 0x4000_0000)").unwrap();
for _ in 0..DDR_FULL_ACCESSIBLE {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_ddr
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// FPGA slave 0 (0x4000_0000 - 0x8000_0000)").unwrap();
for _ in 0..FPGA_SLAVE {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_fpga_slaves
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// FPGA slave 1 (0x8000_0000 - 0xC000_0000)").unwrap();
for _ in 0..FPGA_SLAVE {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_fpga_slaves
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// Unassigned/Reserved (0xC000_0000 - 0xE000_0000)"
)
.unwrap();
for _ in 0..UNASSIGNED_0 {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_unassigned
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// Segments IO peripherals (0xE000_0000 - 0xE030_0000)"
)
.unwrap();
for _ in 0..IO_PERIPHS {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_shared_dev
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// Unassigned/Reserved (0xE030_0000 - 0xE100_0000)"
)
.unwrap();
for _ in 0..UNASSIGNED_1 {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_unassigned
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// NAND (0xE100_0000 - 0xE200_0000)").unwrap();
for _ in 0..NAND {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_shared_dev
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// NOR (0xE200_0000 - 0xE400_0000)").unwrap();
for _ in 0..NOR {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_shared_dev
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// SRAM (0xE400_0000 - 0xE600_0000)").unwrap();
for _ in 0..SRAM {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_sram
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// Unassigned/Reserved (0xE600_0000 - 0xF800_0000)"
)
.unwrap();
for _ in 0..SEGMENTS_UNASSIGNED_2 {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_unassigned
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// AMBA APB peripherals (0xF800_0000 - 0xF900_0000)"
)
.unwrap();
for _ in 0..AMBA_APB {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_shared_dev
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// Unassigned/Reserved (0xF900_0000 - 0xFC00_0000)"
)
.unwrap();
for _ in 0..UNASSIGNED_3 {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_unassigned
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// QSPI XIP (0xFC00_0000 - 0xFE00_0000)").unwrap();
for _ in 0..QSPI_XIP {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_qspi
)
.unwrap();
offset += ONE_MB;
}
writeln!(
buf_writer,
"// Unassiged/Reserved (0xFE00_0000 - 0xFFF0_0000)"
)
.unwrap();
for _ in 0..UNASSIGNED_4 {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_unassigned
)
.unwrap();
offset += ONE_MB;
}
writeln!(buf_writer, "// OCM High (0xFFF0_0000 - 0xFFFF_FFFF)").unwrap();
let mut offset_u64 = offset as u64;
for _ in 0..OCM_MAPPED_HIGH {
writeln!(
buf_writer,
"L1Section::new({}, {}).raw_value(),",
offset, attr_ocm_high
)
.unwrap();
offset_u64 += ONE_MB as u64;
}
// Check that the full 4 GB were covered (not too much, or less)
assert_eq!(offset_u64, 0x1_0000_0000 as u64);
writeln!(buf_writer, "]);").unwrap();
// Finish the file.
drop(buf_writer);
println!("Generated mmu_table.rs");
// Run rustfmt on the generated file
let output = Command::new("rustfmt")
.arg(file_path)
.output()
.expect("Failed to run rustfmt");
if !output.status.success() {
eprintln!("rustfmt failed: {:?}", output);
}
}

11
zynq7000-rt/src/lib.rs Normal file
View File

@ -0,0 +1,11 @@
//! Rust bare metal run-time support for the AMD Zynq 7000 SoCs
//!
//! This includes basic low-level startup code similar to the bare-metal boot routines
//! [provided by Xilinx](https://github.com/Xilinx/embeddedsw/tree/master/lib/bsp/standalone/src/arm/cortexa9/gcc).
#![no_std]
pub mod mmu;
#[cfg(feature = "rt")]
mod mmu_table;
#[cfg(feature = "rt")]
pub mod rt;

191
zynq7000-rt/src/mmu.rs Normal file
View File

@ -0,0 +1,191 @@
//! The overview of translation table memory attributes is described below.
//!
//!| | Memory Range | Definition in Translation Table |
//!|-----------------------|-------------------------|-----------------------------------|
//!| DDR | 0x00000000 - 0x3FFFFFFF | Normal write-back Cacheable |
//!| PL | 0x40000000 - 0xBFFFFFFF | Strongly Ordered |
//!| Reserved | 0xC0000000 - 0xDFFFFFFF | Unassigned |
//!| Memory mapped devices | 0xE0000000 - 0xE02FFFFF | Device Memory |
//!| Reserved | 0xE0300000 - 0xE0FFFFFF | Unassigned |
//!| NAND, NOR | 0xE1000000 - 0xE3FFFFFF | Device memory |
//!| SRAM | 0xE4000000 - 0xE5FFFFFF | Normal write-back Cacheable |
//!| Reserved | 0xE6000000 - 0xF7FFFFFF | Unassigned |
//!| AMBA APB Peripherals | 0xF8000000 - 0xF8FFFFFF | Device Memory |
//!| Reserved | 0xF9000000 - 0xFBFFFFFF | Unassigned |
//!| Linear QSPI - XIP | 0xFC000000 - 0xFDFFFFFF | Normal write-through cacheable |
//!| Reserved | 0xFE000000 - 0xFFEFFFFF | Unassigned |
//!| OCM | 0xFFF00000 - 0xFFFFFFFF | Normal inner write-back cacheable |
//!
//! For region 0x00000000 - 0x3FFFFFFF, a system where DDR is less than 1 GB,
//! region after DDR and before PL is marked as undefined/reserved in translation
//! table. In 0xF8000000 - 0xF8FFFFFF, 0xF8000C00 - 0xF8000FFF, 0xF8010000 -
//! 0xF88FFFFF and 0xF8F03000 to 0xF8FFFFFF are reserved but due to granual size
//! of 1 MB, it is not possible to define separate regions for them. For region
//! 0xFFF00000 - 0xFFFFFFFF, 0xFFF00000 to 0xFFFB0000 is reserved but due to 1MB
//! granual size, it is not possible to define separate region for it.
pub const MAX_DDR_SIZE: usize = 0x4000_0000;
pub const ONE_MB: usize = 0x10_0000;
pub mod offsets {
pub const OFFSET_DDR: usize = 0;
pub const OFFSET_DDR_ALL_ACCESSIBLE: usize = 0x10_0000;
pub const OFFSET_FPGA_SLAVE_0: usize = 0x4000_0000;
pub const OFFSET_FPGA_SLAVE_1_START: usize = 0x8000_0000;
pub const OFFSET_FPGA_SLAVE_1_END: usize = 0xC000_0000;
pub const OFFSET_IO_PERIPHERALS_START: usize = 0xE000_0000;
pub const OFFSET_IO_PERIPHERALS_END: usize = 0xE030_0000;
pub const OFFSET_NAND_MEMORY: usize = 0xE100_0000;
pub const OFFSET_NOR_MEMORY: usize = 0xE200_0000;
pub const OFFSET_SRAM_MEMORY: usize = 0xE400_0000;
pub const OFFSET_SMC_MEMORIES_END: usize = 0xE600_0000;
/// 0xf8000c00 to 0xf8000fff, 0xf8010000 to 0xf88fffff and
/// 0xf8f03000 to 0xf8ffffff are reserved but due to granual size of
/// 1MB, it is not possible to define separate regions for them.
pub const OFFSET_AMBA_APB_START: usize = 0xF800_0000;
pub const OFFSET_AMBA_APB_END: usize = 0xF900_0000;
pub const OFFSET_QSPI_XIP_START: usize = 0xFC00_0000;
pub const OFFSET_QSPI_XIP_END: usize = 0xFE00_0000;
/// 0xfff00000 to 0xfffb0000 is reserved but due to granual size of
/// 1MB, it is not possible to define separate region for it
pub const OFFSET_OCM_MAPPED_HIGH_START: usize = 0xFFF0_0000;
pub const OFFSET_OCM_MAPPED_HIGH_END: u64 = 0x1_0000_0000;
}
pub mod segments {
pub use super::offsets::*;
use super::{MAX_DDR_SIZE, ONE_MB};
/// First 1 MB of DDR has special treatment, access is dependant on SCU/OCM state.
/// Refer to Zynq TRM UG585 p.106 for more details.
pub const DDR_FULL_ACCESSIBLE: usize = (MAX_DDR_SIZE - ONE_MB) / ONE_MB;
pub const FPGA_SLAVE: usize = (OFFSET_FPGA_SLAVE_1_START - OFFSET_FPGA_SLAVE_0) / ONE_MB;
pub const UNASSIGNED_0: usize =
(OFFSET_IO_PERIPHERALS_START - OFFSET_FPGA_SLAVE_1_END) / ONE_MB;
pub const IO_PERIPHS: usize =
(OFFSET_IO_PERIPHERALS_END - OFFSET_IO_PERIPHERALS_START) / ONE_MB;
pub const UNASSIGNED_1: usize = (OFFSET_NAND_MEMORY - OFFSET_IO_PERIPHERALS_END) / ONE_MB;
pub const NAND: usize = (OFFSET_NOR_MEMORY - OFFSET_NAND_MEMORY) / ONE_MB;
pub const NOR: usize = (OFFSET_SRAM_MEMORY - OFFSET_NOR_MEMORY) / ONE_MB;
pub const SRAM: usize = (OFFSET_SMC_MEMORIES_END - OFFSET_SRAM_MEMORY) / ONE_MB;
pub const SEGMENTS_UNASSIGNED_2: usize =
(OFFSET_AMBA_APB_START - OFFSET_SMC_MEMORIES_END) / ONE_MB;
pub const AMBA_APB: usize = (OFFSET_AMBA_APB_END - OFFSET_AMBA_APB_START) / ONE_MB;
pub const UNASSIGNED_3: usize = (OFFSET_QSPI_XIP_START - OFFSET_AMBA_APB_END) / ONE_MB;
pub const QSPI_XIP: usize = (OFFSET_QSPI_XIP_END - OFFSET_QSPI_XIP_START) / ONE_MB;
pub const UNASSIGNED_4: usize = (OFFSET_OCM_MAPPED_HIGH_START - OFFSET_QSPI_XIP_END) / ONE_MB;
pub const OCM_MAPPED_HIGH: usize = ((OFFSET_OCM_MAPPED_HIGH_END
- OFFSET_OCM_MAPPED_HIGH_START as u64)
/ ONE_MB as u64) as usize;
}
pub mod section_attrs {
use cortex_ar::mmu::{
AccessPermissions, CacheableMemoryAttribute, MemoryRegionAttributes, SectionAttributes,
};
pub const DDR: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: true,
access: AccessPermissions::FullAccess,
// Manager domain
domain: 0b1111,
execute_never: false,
memory_attrs: MemoryRegionAttributes::CacheableMemory {
inner: CacheableMemoryAttribute::WriteBackWriteAlloc,
outer: CacheableMemoryAttribute::WriteBackWriteAlloc,
}
.as_raw(),
};
pub const FPGA_SLAVES: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: false,
access: AccessPermissions::FullAccess,
domain: 0b0000,
execute_never: false,
memory_attrs: MemoryRegionAttributes::StronglyOrdered.as_raw(),
};
pub const SHAREABLE_DEVICE: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: false,
access: AccessPermissions::FullAccess,
domain: 0b0000,
execute_never: false,
memory_attrs: MemoryRegionAttributes::ShareableDevice.as_raw(),
};
pub const SRAM: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: false,
access: AccessPermissions::FullAccess,
domain: 0b0000,
execute_never: false,
memory_attrs: MemoryRegionAttributes::OuterAndInnerWriteBackNoWriteAlloc.as_raw(),
};
pub const QSPI_XIP: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: false,
access: AccessPermissions::FullAccess,
domain: 0b0000,
execute_never: false,
memory_attrs: MemoryRegionAttributes::OuterAndInnerWriteThroughNoWriteAlloc.as_raw(),
};
pub const OCM_MAPPED_HIGH: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: false,
access: AccessPermissions::FullAccess,
domain: 0b0000,
execute_never: false,
memory_attrs: MemoryRegionAttributes::CacheableMemory {
inner: CacheableMemoryAttribute::WriteThroughNoWriteAlloc,
outer: CacheableMemoryAttribute::NonCacheable,
}
.as_raw(),
};
pub const UNASSIGNED_RESERVED: SectionAttributes = SectionAttributes {
non_global: false,
p_bit: false,
shareable: false,
access: AccessPermissions::PermissionFault,
domain: 0b0000,
execute_never: false,
memory_attrs: MemoryRegionAttributes::StronglyOrdered.as_raw(),
};
}
pub const NUM_L1_PAGE_TABLE_ENTRIES: usize = 4096;
#[repr(C, align(16384))]
#[cfg(feature = "rt")]
pub struct L1Table(pub(crate) [u32; NUM_L1_PAGE_TABLE_ENTRIES]);
/// Load the MMU translation table base address into the MMU.
///
/// # Safety
///
/// This function is unsafe because it directly writes to the MMU related registers. It has to be
/// called once in the boot code before enabling the MMU, and it should be called while the MMU is
/// disabled.
#[unsafe(no_mangle)]
#[cfg(feature = "rt")]
unsafe extern "C" fn load_mmu_table() {
let table_base = &crate::mmu_table::MMU_L1_PAGE_TABLE.0 as *const _ as u32;
unsafe {
core::arch::asm!(
"orr {0}, {0}, #0x5B", // Outer-cacheable, WB
"mcr p15, 0, {0}, c2, c0, 0", // Load table pointer
inout(reg) table_base => _,
options(nostack, preserves_flags)
);
}
}

4121
zynq7000-rt/src/mmu_table.rs Normal file
View File

File diff suppressed because it is too large Load Diff

396
zynq7000-rt/src/rt.rs Normal file
View File

@ -0,0 +1,396 @@
//! Start-up code for Zynq 7000
//!
//! The bootup routine was kepts as similar to the one
//! [provided by Xilinx](https://github.com/Xilinx/embeddedsw/blob/master/lib/bsp/standalone/src/arm/cortexa9/gcc/boot.S)
//! as possible. The boot routine includes stack, MMU, cache and .bss/.data section initialization.
use cortex_a_rt as _;
use cortex_ar::register::{Cpsr, cpsr::ProcessorMode};
// Start-up code for Armv7-A
//
// We set up our stacks and `kmain` in system mode.
core::arch::global_asm!(
r#"
.set PSS_L2CC_BASE_ADDR, 0xF8F02000
.set PSS_SLCR_BASE_ADDR, 0xF8000000
.set RESERVED, 0x0fffff00
.set LRemap, 0xFE00000F /* set the base address of the peripheral block as not shared */
.set L2CCWay, (PSS_L2CC_BASE_ADDR + 0x077C) /*(PSS_L2CC_BASE_ADDR + PSS_L2CC_CACHE_INVLD_WAY_OFFSET)*/
.set L2CCSync, (PSS_L2CC_BASE_ADDR + 0x0730) /*(PSS_L2CC_BASE_ADDR + PSS_L2CC_CACHE_SYNC_OFFSET)*/
.set L2CCCrtl, (PSS_L2CC_BASE_ADDR + 0x0100) /*(PSS_L2CC_BASE_ADDR + PSS_L2CC_CNTRL_OFFSET)*/
.set L2CCAuxCrtl, (PSS_L2CC_BASE_ADDR + 0x0104) /*(PSS_L2CC_BASE_ADDR + XPSS_L2CC_AUX_CNTRL_OFFSET)*/
.set L2CCTAGLatReg, (PSS_L2CC_BASE_ADDR + 0x0108) /*(PSS_L2CC_BASE_ADDR + XPSS_L2CC_TAG_RAM_CNTRL_OFFSET)*/
.set L2CCDataLatReg, (PSS_L2CC_BASE_ADDR + 0x010C) /*(PSS_L2CC_BASE_ADDR + XPSS_L2CC_DATA_RAM_CNTRL_OFFSET)*/
.set L2CCIntClear, (PSS_L2CC_BASE_ADDR + 0x0220) /*(PSS_L2CC_BASE_ADDR + XPSS_L2CC_IAR_OFFSET)*/
.set L2CCIntRaw, (PSS_L2CC_BASE_ADDR + 0x021C) /*(PSS_L2CC_BASE_ADDR + XPSS_L2CC_ISR_OFFSET)*/
.set SLCRlockReg, (PSS_SLCR_BASE_ADDR + 0x04) /*(PSS_SLCR_BASE_ADDR + XPSS_SLCR_LOCK_OFFSET)*/
.set SLCRUnlockReg, (PSS_SLCR_BASE_ADDR + 0x08) /*(PSS_SLCR_BASE_ADDR + XPSS_SLCR_UNLOCK_OFFSET)*/
.set SLCRL2cRamReg, (PSS_SLCR_BASE_ADDR + 0xA1C) /*(PSS_SLCR_BASE_ADDR + XPSS_SLCR_L2C_RAM_OFFSET)*/
.set SLCRCPURSTReg, (0xF8000000 + 0x244) /*(XPS_SYS_CTRL_BASEADDR + A9_CPU_RST_CTRL_OFFSET)*/
.set EFUSEStatus, (0xF800D000 + 0x10) /*(XPS_EFUSE_BASEADDR + EFUSE_STATUS_OFFSET)*/
.set CRValMmuCac, 0b01000000000101 /* Enable IDC, and MMU */
.set CRValHiVectorAddr, 0b10000000000000 /* Set the Vector address to high, 0xFFFF0000 */
.set L2CCAuxControl, 0x72360000 /* Enable all prefetching, Cache replacement policy, Parity enable,
Event monitor bus enable and Way Size (64 KB) */
.set L2CCControl, 0x01 /* Enable L2CC */
.set L2CCTAGLatency, 0x0111 /* latency for TAG RAM */
.set L2CCDataLatency, 0x0121 /* latency for DATA RAM */
.set SLCRlockKey, 0x767B /* SLCR lock key */
.set SLCRUnlockKey, 0xDF0D /* SLCR unlock key */
.set SLCRL2cRamConfig, 0x00020202 /* SLCR L2C ram configuration */
.set FPEXC_EN, 0x40000000 /* FPU enable bit, (1 << 30) */
.section .text.startup
.align 0
.global _start
.type _start, %function
_start:
// only allow cpu0 through
// Read MPIDR
mrc p15,0,r1,c0,c0,5
// Extract CPU ID bits. For single-core systems, this should always be 0
and r1, r1, #0x3
cmp r1, #0
beq check_efuse
b initialize
// Zynq specific code. It is recommended to reset CPU1 according to page 160 of the datasheet
check_efuse:
ldr r0,=EFUSEStatus
ldr r1,[r0] /* Read eFuse setting */
ands r1,r1,#0x80 /* Check whether device is having single core */
beq initialize
/* single core device, reset cpu1 */
ldr r0,=SLCRUnlockReg /* Load SLCR base address base + unlock register */
ldr r1,=SLCRUnlockKey /* set unlock key */
str r1, [r0] /* Unlock SLCR */
ldr r0,=SLCRCPURSTReg
ldr r1,[r0] /* Read CPU Software Reset Control register */
orr r1,r1,#0x22
str r1,[r0] /* Reset CPU1 */
ldr r0,=SLCRlockReg /* Load SLCR base address base + lock register */
ldr r1,=SLCRlockKey /* set lock key */
str r1, [r0] /* lock SLCR */
initialize:
mrc p15, 0, r0, c0, c0, 0 /* Get the revision */
and r5, r0, #0x00f00000
and r6, r0, #0x0000000f
orr r6, r6, r5, lsr #20-4
/* set VBAR to the _vector_table address in linker script */
ldr r0, =_vector_table
mcr p15, 0, r0, c12, c0, 0
/* Invalidate scu */
ldr r7, =0xf8f0000c
ldr r6, =0xffff
str r6, [r7]
/* Invalidate caches and TLBs */
mov r0,#0 /* r0 = 0 */
mcr p15, 0, r0, c8, c7, 0 /* invalidate TLBs */
mcr p15, 0, r0, c7, c5, 0 /* invalidate icache */
mcr p15, 0, r0, c7, c5, 6 /* Invalidate branch predictor array */
bl invalidate_dcache /* invalidate dcache */
/* Disable MMU, if enabled */
mrc p15, 0, r0, c1, c0, 0 /* read CP15 register 1 */
bic r0, r0, #0x1 /* clear bit 0 */
mcr p15, 0, r0, c1, c0, 0 /* write value back */
// Set up stacks first.
ldr r3, =_stack_top
// get the current PSR
mrs r0, cpsr
// mask for mode bits
mvn r1, #0x1f
and r2, r1, r0
// IRQ mode
orr r2, r2, {irq_mode}
msr cpsr, r2
// IRQ stack pointer
mov sp, r3
ldr r1, =_irq_stack_size
sub r3, r3, r1
mrs r0, cpsr
and r2, r1, r0
// Supervisor mode
orr r2, r2, {svc_mode}
msr cpsr, r2
// Supervisor stack pointer
mov sp, r3
ldr r1, =_svc_stack_size
sub r3, r3, r1
mrs r0, cpsr
and r2, r1, r0
// Abort mode
orr r2, r2, {abt_mode}
msr cpsr, r2
// Abort stack pointer
mov sp, r3
ldr r1, =_abt_stack_size
sub r3, r3, r1
mrs r0, cpsr
and r2, r1, r0
// FIQ mode
orr r2, r2, {fiq_mode}
msr cpsr, r2
// FIQ stack pointer
mov sp, r3
ldr r1, =_fiq_stack_size
sub r3, r3, r1
mrs r0, cpsr
and r2, r1, r0
// Undefined mode
orr r2, r2, {und_mode}
msr cpsr, r2
// Undefined stack pointer
mov sp, r3
ldr r1, =_und_stack_size
sub r3, r3, r1
mrs r0, cpsr
and r2, r1, r0
// System mode
orr r2, r2, {sys_mode}
msr cpsr, r2
// System stack pointer (main stack)
mov sp, r3
// set scu enable bit in scu
ldr r7, =0xf8f00000
ldr r0, [r7]
orr r0, r0, #0x1
str r0, [r7]
/* enable MMU and cache */
bl load_mmu_table
mvn r0,#0 /* Load MMU domains -- all ones=manager */
mcr p15,0,r0,c3,c0,0
/* Enable mmu, icahce and dcache */
ldr r0,=CRValMmuCac
mcr p15,0,r0,c1,c0,0 /* Enable cache and MMU */
dsb /* dsb allow the MMU to start up */
isb /* isb flush prefetch buffer */
/* Write to ACTLR */
mrc p15, 0, r0, c1, c0, 1 /* Read ACTLR*/
orr r0, r0, #(0x01 << 6) /* set SMP bit */
orr r0, r0, #(0x01 ) /* Cache/TLB maintenance broadcast */
mcr p15, 0, r0, c1, c0, 1 /* Write ACTLR*/
/* Invalidate L2 Cache and enable L2 Cache*/
/* For AMP, assume running on CPU1. Don't initialize L2 Cache (up to Linux) */
ldr r0,=L2CCCrtl /* Load L2CC base address base + control register */
mov r1, #0 /* force the disable bit */
str r1, [r0] /* disable the L2 Caches */
ldr r0,=L2CCAuxCrtl /* Load L2CC base address base + Aux control register */
ldr r1,[r0] /* read the register */
ldr r2,=L2CCAuxControl /* set the default bits */
orr r1,r1,r2
str r1, [r0] /* store the Aux Control Register */
ldr r0,=L2CCTAGLatReg /* Load L2CC base address base + TAG Latency address */
ldr r1,=L2CCTAGLatency /* set the latencies for the TAG*/
str r1, [r0] /* store the TAG Latency register Register */
ldr r0,=L2CCDataLatReg /* Load L2CC base address base + Data Latency address */
ldr r1,=L2CCDataLatency /* set the latencies for the Data*/
str r1, [r0] /* store the Data Latency register Register */
ldr r0,=L2CCWay /* Load L2CC base address base + way register*/
ldr r2, =0xFFFF
str r2, [r0] /* force invalidate */
ldr r0,=L2CCSync /* need to poll 0x730, PSS_L2CC_CACHE_SYNC_OFFSET */
/* Load L2CC base address base + sync register*/
/* poll for completion */
Sync:
ldr r1, [r0]
cmp r1, #0
bne Sync
ldr r0,=L2CCIntRaw /* clear pending interrupts */
ldr r1,[r0]
ldr r0,=L2CCIntClear
str r1,[r0]
ldr r0,=SLCRUnlockReg /* Load SLCR base address base + unlock register */
ldr r1,=SLCRUnlockKey /* set unlock key */
str r1, [r0] /* Unlock SLCR */
ldr r0,=SLCRL2cRamReg /* Load SLCR base address base + l2c Ram Control register */
ldr r1,=SLCRL2cRamConfig /* set the configuration value */
str r1, [r0] /* store the L2c Ram Control Register */
ldr r0,=SLCRlockReg /* Load SLCR base address base + lock register */
ldr r1,=SLCRlockKey /* set lock key */
str r1, [r0] /* lock SLCR */
ldr r0,=L2CCCrtl /* Load L2CC base address base + control register */
ldr r1,[r0] /* read the register */
mov r2, #L2CCControl /* set the enable bit */
orr r1,r1,r2
str r1, [r0] /* enable the L2 Caches */
mov r0, r0
mrc p15, 0, r1, c1, c0, 2 /* read cp access control register (CACR) into r1 */
orr r1, r1, #(0xf << 20) /* enable full access for p10 & p11 */
mcr p15, 0, r1, c1, c0, 2 /* write back into CACR */
/* enable vfp */
fmrx r1, FPEXC /* read the exception register */
orr r1,r1, #FPEXC_EN /* set VFP enable bit, leave the others in orig state */
fmxr FPEXC, r1 /* write back the exception register */
mrc p15,0,r0,c1,c0,0 /* flow prediction enable */
orr r0, r0, #(0x01 << 11) /* #0x8000 */
mcr p15,0,r0,c1,c0,0
mrc p15,0,r0,c1,c0,1 /* read Auxiliary Control Register */
orr r0, r0, #(0x1 << 2) /* enable Dside prefetch */
orr r0, r0, #(0x1 << 1) /* enable L2 Prefetch hint */
mcr p15,0,r0,c1,c0,1 /* write Auxiliary Control Register */
mrs r0, cpsr /* get the current PSR */
bic r0, r0, #0x100 /* enable asynchronous abort exception */
msr cpsr_xsf, r0
/* Zero BSS and initialize data before calling any function which might require them. */
// Initialise .bss
ldr r0, =__sbss
ldr r1, =__ebss
mov r2, 0
0:
cmp r1, r0
beq 1f
stm r0!, {{r2}}
b 0b
1:
// Initialise .data
ldr r0, =__sdata
ldr r1, =__edata
ldr r2, =__sidata
0:
cmp r1, r0
beq 1f
ldm r2!, {{r3}}
stm r0!, {{r3}}
b 0b
1:
// Jump to application
// Load CPU ID 0, which will be used as a function argument to the boot_core function.
mov r0, #0x0
bl boot_core
// In case the application returns, loop forever
b .
.size _start, . - _start
.type _invalidate_dcache, %function
invalidate_dcache:
mrc p15, 1, r0, c0, c0, 1 /* read CLIDR */
ands r3, r0, #0x7000000
mov r3, r3, lsr #23 /* cache level value (naturally aligned) */
beq finished
mov r10, #0 /* start with level 0 */
loop1:
add r2, r10, r10, lsr #1 /* work out 3xcachelevel */
mov r1, r0, lsr r2 /* bottom 3 bits are the Cache type for this level */
and r1, r1, #7 /* get those 3 bits alone */
cmp r1, #2
blt skip /* no cache or only instruction cache at this level */
mcr p15, 2, r10, c0, c0, 0 /* write the Cache Size selection register */
isb /* isb to sync the change to the CacheSizeID reg */
mrc p15, 1, r1, c0, c0, 0 /* reads current Cache Size ID register */
and r2, r1, #7 /* extract the line length field */
add r2, r2, #4 /* add 4 for the line length offset (log2 16 bytes) */
ldr r4, =0x3ff
ands r4, r4, r1, lsr #3 /* r4 is the max number on the way size (right aligned) */
clz r5, r4 /* r5 is the bit position of the way size increment */
ldr r7, =0x7fff
ands r7, r7, r1, lsr #13 /* r7 is the max number of the index size (right aligned) */
loop2:
mov r9, r4 /* r9 working copy of the max way size (right aligned) */
loop3:
orr r11, r10, r9, lsl r5 /* factor in the way number and cache number into r11 */
orr r11, r11, r7, lsl r2 /* factor in the index number */
mcr p15, 0, r11, c7, c6, 2 /* invalidate by set/way */
subs r9, r9, #1 /* decrement the way number */
bge loop3
subs r7, r7, #1 /* decrement the index */
bge loop2
skip:
add r10, r10, #2 /* increment the cache number */
cmp r3, r10
bgt loop1
finished:
mov r10, #0 /* switch back to cache level 0 */
mcr p15, 2, r10, c0, c0, 0 /* select current cache level in cssr */
dsb
isb
bx lr
.size invalidate_dcache, . - invalidate_dcache
"#,
fiq_mode = const {
Cpsr::new_with_raw_value(0)
.with_mode(ProcessorMode::Fiq)
.with_i(true)
.with_f(true)
.raw_value()
},
irq_mode = const {
Cpsr::new_with_raw_value(0)
.with_mode(ProcessorMode::Irq)
.with_i(true)
.with_f(true)
.raw_value()
},
svc_mode = const {
Cpsr::new_with_raw_value(0)
.with_mode(ProcessorMode::Svc)
.with_i(true)
.with_f(true)
.raw_value()
},
und_mode = const {
Cpsr::new_with_raw_value(0)
.with_mode(ProcessorMode::Und)
.with_i(true)
.with_f(true)
.raw_value()
},
abt_mode = const {
Cpsr::new_with_raw_value(0)
.with_mode(ProcessorMode::Abt)
.with_i(true)
.with_f(true)
.raw_value()
},
sys_mode = const {
Cpsr::new_with_raw_value(0)
.with_mode(ProcessorMode::Sys)
.with_i(true)
.with_f(true)
.raw_value()
},
);