From cb466c24a3d001ddd370744ea04e9b75a8f33758 Mon Sep 17 00:00:00 2001 From: bunnie Date: Tue, 1 Oct 2024 14:39:40 +0800 Subject: [PATCH] recognize more flash chip types 1. System works with 1.8V VDD I/O now, needs a new chip ID for flash 2. Add direct c-translated version of init_clock_asic() for reference testing. --- loader/src/platform/cramium/cramium.rs | 190 +++++++++++++++++++++++-- 1 file changed, 180 insertions(+), 10 deletions(-) diff --git a/loader/src/platform/cramium/cramium.rs b/loader/src/platform/cramium/cramium.rs index d7baff115..db57464dc 100644 --- a/loader/src/platform/cramium/cramium.rs +++ b/loader/src/platform/cramium/cramium.rs @@ -140,6 +140,7 @@ pub fn early_init() { // Now setup the clocks for real // Safety: this can only be called in the early_init boot context let perclk = unsafe { init_clock_asic(800_000_000) }; + // let perclk = unsafe { init_clock_asic_c(800_000_000, 0) }; crate::println!("Perclk is {} Hz", perclk); // Configure the UDMA UART. This UART's settings will be used as the initial console UART. @@ -147,10 +148,10 @@ pub fn early_init() { // on the cramium-hal crate to be functional. // Set up the IO mux to map UART_A0: - // UART_RX_A[0] = PA3 - // UART_TX_A[0] = PA4 - // UART_RX_A[1] = PD13 - // UART_RX_A[1] = PD14 + // UART_RX_A[0] = PA3 app + // UART_TX_A[0] = PA4 app + // UART_RX_A[1] = PD13 console + // UART_RX_A[1] = PD14 console let mut iox = Iox::new(utra::iox::HW_IOX_BASE as *mut u32); iox.set_alternate_function(IoxPort::PD, 13, IoxFunction::AF1); iox.set_alternate_function(IoxPort::PD, 14, IoxFunction::AF1); @@ -393,7 +394,7 @@ pub fn early_init() { 11, Some(IoxFunction::AF1), Some(IoxDir::Output), - Some(IoxDriveStrength::Drive4mA), + Some(IoxDriveStrength::Drive2mA), Some(IoxEnable::Enable), None, None, @@ -407,7 +408,7 @@ pub fn early_init() { Some(IoxFunction::AF1), None, Some(IoxDriveStrength::Drive2mA), - Some(IoxEnable::Enable), + Some(IoxEnable::Disable), None, None, ); @@ -420,7 +421,7 @@ pub fn early_init() { Some(IoxFunction::AF1), Some(IoxDir::Output), Some(IoxDriveStrength::Drive2mA), - Some(IoxEnable::Enable), + Some(IoxEnable::Disable), None, None, ); @@ -432,7 +433,7 @@ pub fn early_init() { Some(IoxFunction::AF1), Some(IoxDir::Output), Some(IoxDriveStrength::Drive2mA), - Some(IoxEnable::Enable), + Some(IoxEnable::Disable), None, None, ); @@ -507,7 +508,8 @@ pub fn early_init() { crate::println!("flash ID: {:x}", flash_id); crate::println!("ram ID: {:x}", ram_id); // density 18, memory type 20, mfg ID C2 ==> MX25L128833F - assert!(flash_id & 0xFF_FF_FF == 0x1820C2); + // density 38, memory type 25, mfg ID C2 ==> MX25U12832F + assert!(flash_id & 0xFF_FF_FF == 0x1820C2 || flash_id & 0xFF_FF_FF == 0x38_25_C2); // KGD 5D, mfg ID 9D; remainder of bits are part of the EID assert!(ram_id & 0xFF_FF == 0x5D9D); @@ -534,7 +536,8 @@ pub fn early_init() { crate::println!("QPI flash ID: {:x}", flash_id); crate::println!("QPI ram ID: {:x}", ram_id); // density 18, memory type 20, mfg ID C2 ==> MX25L128833F - assert!(flash_id & 0xFF_FF_FF == 0x1820C2); + // density 38, memory type 25, mfg ID C2 ==> MX25U12832F + assert!(flash_id & 0xFF_FF_FF == 0x1820C2 || flash_id & 0xFF_FF_FF == 0x38_25_C2); // KGD 5D, mfg ID 9D; remainder of bits are part of the EID assert!(ram_id & 0xFF_FF == 0x5D9D); @@ -786,6 +789,7 @@ pub fn early_init() { } } + /* let aoc_base = utralib::CSR::new(0x4006_0000 as *mut u32); unsafe { for i in 0..0x50 / 4 { @@ -796,6 +800,7 @@ pub fn early_init() { aoc_base.base().add(4).write_volatile(pmu_cr & !1); crate::println!("pmu_cr upd: {:x}", aoc_base.base().add(4).read_volatile()); } + */ udma_uart.write("Press any key to continue...".as_bytes()); getc(); udma_uart.write(b"\n\rBooting!\n\r"); @@ -1361,3 +1366,168 @@ pub extern "C" fn trap_handler( unsafe { mie::set_mext() }; unsafe { _resume_context(0x61008000u32) }; // this is the scratch page used in the assembly routine above } + +#[allow(dead_code)] +pub fn log_2(mut value: u32) -> u32 { + let mut result = 0; + + // Shift right until we find the position of the highest set bit + while value > 1 { + value >>= 1; + result += 1; + } + + result +} + +#[allow(dead_code)] +/// Direct translation of the C code +pub unsafe fn init_clock_asic_c(freq_hz: u32, duty_sram: u32) -> u32 { + use utra::sysctrl; + let daric_cgu = sysctrl::HW_SYSCTRL_BASE as *mut u32; + + const UNIT_MHZ: u32 = 1000u32 * 1000u32; + const PFD_F_MHZ: u32 = 16; + const FREQ_0: u32 = 16u32 * UNIT_MHZ; + const FREQ_OSC_MHZ: u32 = 48; // Actually 48MHz + const M: u32 = FREQ_OSC_MHZ / PFD_F_MHZ; // - 1; // OSC input was 24, replace with 48 + + const TBL_Q: [u16; 7] = [ + // keep later DIV even number as possible + 0x7777, // 16-32 MHz + 0x7737, // 32-64 + 0x3733, // 64-128 + 0x3313, // 128-256 + 0x3311, // 256-512 // keep ~ 100MHz + 0x3301, // 512-1024 + 0x3301, // 1024-1600 + ]; + const TBL_MUL: [u32; 7] = [ + 64, // 16-32 MHz + 32, // 32-64 + 16, // 64-128 + 8, // 128-256 + 4, // 256-512 + 2, // 512-1024 + 2, // 1024-1600 + ]; + + if (0 == (daric_cgu.add(sysctrl::SFR_IPCPLLMN.offset()).read_volatile() & 0x0001F000)) + || (0 == (daric_cgu.add(sysctrl::SFR_IPCPLLMN.offset()).read_volatile() & 0x00000fff)) + { + // for SIM, avoid div by 0 if unconfigurated + // , default VCO 48MHz / 48 * 1200 = 1.2GHz + // TODO magic numbers + daric_cgu + .add(sysctrl::SFR_IPCPLLMN.offset()) + .write_volatile(((M << 12) & 0x0001F000) | ((1200) & 0x00000fff)); + daric_cgu.add(sysctrl::SFR_IPCPLLF.offset()).write_volatile(0); // ?? + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_IPCARIPFLOW.offset()).write_volatile(0x32); // commit + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + } + + // TODO select int/ext osc/xtal + daric_cgu.add(sysctrl::SFR_CGUSEL1.offset()).write_volatile(1); // 0: RC, 1: XTAL + daric_cgu.add(sysctrl::SFR_CGUFSCR.offset()).write_volatile(FREQ_OSC_MHZ); // external crystal is 48MHz + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_CGUSET.offset()).write_volatile(0x32); + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + if freq_hz < 1000000 { + daric_cgu.add(sysctrl::SFR_IPCOSC.offset()).write_volatile(freq_hz); + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_IPCARIPFLOW.offset()).write_volatile(0x32); // commit, must write 32 + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + } + // switch to OSC + daric_cgu.add(sysctrl::SFR_CGUSEL0.offset()).write_volatile(0); // clktop sel, 0:clksys, 1:clkpll0 + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_CGUSET.offset()).write_volatile(0x32); // commit + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + if freq_hz < 1000000 { + } else { + let f16_mhz_log2: usize = log_2(freq_hz / FREQ_0) as usize; + + // PD PLL + daric_cgu + .add(sysctrl::SFR_IPCLPEN.offset()) + .write_volatile(daric_cgu.add(sysctrl::SFR_IPCLPEN.offset()).read_volatile() | 0x02); + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_IPCARIPFLOW.offset()).write_volatile(0x32); // commit, must write 32 + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + // delay + for _ in 0..1024 { + unsafe { core::arch::asm!("nop") }; + } + + let n_fxp24: u64 = (((freq_hz as u64) << 24u64) * TBL_MUL[f16_mhz_log2] as u64 + + PFD_F_MHZ as u64 * UNIT_MHZ as u64 / 2u64) + / (PFD_F_MHZ as u64 * UNIT_MHZ as u64); // rounded + let n_frac: u32 = (n_fxp24 & 0x00ffffff) as u32; + + daric_cgu + .add(sysctrl::SFR_IPCPLLMN.offset()) + .write_volatile(((M << 12) & 0x0001F000) | ((n_fxp24 >> 24) & 0x00000fff) as u32); // 0x1F598; // ?? + daric_cgu + .add(sysctrl::SFR_IPCPLLF.offset()) + .write_volatile(n_frac | if 0 == n_frac { 0 } else { 1 << 24 }); // ?? + daric_cgu.add(sysctrl::SFR_IPCPLLQ.offset()).write_volatile(TBL_Q[f16_mhz_log2] as u32); // ?? TODO select DIV for VCO freq + + // VCO bias CPP bias CPI bias + // 1 2 3 + //DARIC_IPC->ipc = (3 << 6) | (5 << 3) | (5); + daric_cgu.add(sysctrl::SFR_IPCCR.offset()).write_volatile((1 << 6) | (2 << 3) | (3)); + // daric_cgu.add(sysctrl::SFR_IPCCR.offset()).write_volatile((3 << 6) | (5 << 3) | (5)); + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_IPCARIPFLOW.offset()).write_volatile(0x32); // commit, must write 32 + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + daric_cgu + .add(sysctrl::SFR_IPCLPEN.offset()) + .write_volatile(daric_cgu.add(sysctrl::SFR_IPCLPEN.offset()).read_volatile() & !0x02); + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_IPCARIPFLOW.offset()).write_volatile(0x32); // commit, must write 32 + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + // delay + for _ in 0..1024 { + unsafe { core::arch::asm!("nop") }; + } + //printf("read reg a0 : %08" PRIx32"\n", *((volatile uint32_t* )0x400400a0)); + //printf("read reg a4 : %04" PRIx16"\n", *((volatile uint16_t* )0x400400a4)); + //printf("read reg a8 : %04" PRIx16"\n", *((volatile uint16_t* )0x400400a8)); + crate::println!("PLL switchover"); + // TODO wait/poll lock status? + daric_cgu.add(sysctrl::SFR_CGUSEL0.offset()).write_volatile(1); // clktop sel, 0:clksys, 1:clkpll0 + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(sysctrl::SFR_CGUSET.offset()).write_volatile(0x32); // commit + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + crate::println!("PLL switchover done"); + + // printf (" MN: 0x%05x, F: 0x%06x, Q: 0x%04x\n", + // DARIC_IPC->pll_mn, DARIC_IPC->pll_f, DARIC_IPC->pll_q); + // printf (" LPEN: 0x%01x, OSC: 0x%04x, BIAS: 0x%04x,\n", + // DARIC_IPC->lpen, DARIC_IPC->osc, DARIC_IPC->ipc); + } + + daric_cgu.add(utra::sysctrl::SFR_CGUFD_CFGFDCR_0_4_0.offset()).write_volatile(0x7fff); // CPU + if 0 == duty_sram { + daric_cgu.add(utra::sysctrl::SFR_CGUFD_CFGFDCR_0_4_1.offset()).write_volatile(0x3f7f); // aclk + } else { + daric_cgu.add(utra::sysctrl::SFR_CGUFD_CFGFDCR_0_4_1.offset()).write_volatile(duty_sram); + } + daric_cgu.add(utra::sysctrl::SFR_CGUFD_CFGFDCR_0_4_2.offset()).write_volatile(0x1f3f); // hclk + daric_cgu.add(utra::sysctrl::SFR_CGUFD_CFGFDCR_0_4_3.offset()).write_volatile(0x0f1f); // iclk + daric_cgu.add(utra::sysctrl::SFR_CGUFD_CFGFDCR_0_4_4.offset()).write_volatile(0x070f); // pclk + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + daric_cgu.add(utra::sysctrl::SFR_CGUSET.offset()).write_volatile(0x32); // commit + core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + // UDMACORE->CFG_CG = 0xffffffff; //everything on + // core::sync::atomic::compiler_fence(core::sync::atomic::Ordering::SeqCst); + + 100_000_000 // bodge for now +}