diff options
| -rw-r--r-- | build.zig | 11 | ||||
| -rw-r--r-- | experiments/length-CPU360-final.csv | 46 | ||||
| -rw-r--r-- | experiments/length-CPU360-fine.csv | 46 | ||||
| -rw-r--r-- | experiments/length-CPU360.csv | 36 | ||||
| -rw-r--r-- | src/hal/clkrst.zig | 90 | ||||
| -rw-r--r-- | src/pardes/app.zig | 29 | ||||
| -rw-r--r-- | src/soc.zig | 5 | ||||
| -rw-r--r-- | tools/rtt.zig | 11 |
8 files changed, 271 insertions, 3 deletions
@@ -108,6 +108,14 @@ pub fn build(b: *std.Build) void { // cycle counts. Off by default because it puts a line on the wire per frame, which is the very // resource being measured - it answers "where did the 34 ms go", not "how fast is it". options.addOption(bool, "prof", b.option(bool, "prof", "print per-phase cycle counts (pardes)") orelse false); + // The CPU clock, in MHz. The bootloader leaves 90; the CPLL is already at 360, so 180 and 360 + // are a divider change away and nothing else (see hal/clkrst.zig:setCpuFreq). Opt-in rather + // than default because it is the one setting here that changes how every other timing in the + // image behaves, and because 90 is what every measurement in experiments/ was taken against. + const cpu_mhz = b.option(u16, "cpu-mhz", "HP CPU clock: 90 (bootloader default), 180 or 360") orelse 90; + if (cpu_mhz != 90 and cpu_mhz != 180 and cpu_mhz != 360) + std.debug.panic("-Dcpu-mhz must be 90, 180 or 360; the P4's CPLL divides 360 by 4, 2 or 1", .{}); + options.addOption(u16, "cpu_mhz", cpu_mhz); options.addOption([]const u8, "wifi_ssid", wifi_ssid); options.addOption([]const u8, "wifi_psk", if (psk_file) |path| blk: { const raw = std.Io.Dir.cwd().readFileAlloc(b.graph.io, path, b.allocator, .limited(256)) catch @@ -1216,6 +1224,9 @@ fn linkerScript(b: *std.Build, stack_size: u32, peripherals_ld: ?[]const u8) []c \\/* ESP32-P4 mask ROM entry points, the only "library" this image links against */ \\ets_printf = 0x4fc00024; \\ets_delay_us = 0x4fc0003c; + \\/* Recalibrate the ROM's cycles-per-microsecond after a CPU clock change; without it every + \\ ets_delay_us is wrong by exactly the frequency ratio (esp32p4.rom.ld:32). */ + \\ets_update_cpu_frequency = 0x4fc00044; \\/* Invalidate the caches. The bootloader leaves lines that do not match the mapping it \\ finally installs, so a large image reads its own .rodata and gets its own .text back. */ \\Cache_Invalidate_All = 0x4fc00404; diff --git a/experiments/length-CPU360-final.csv b/experiments/length-CPU360-final.csv new file mode 100644 index 0000000..b7505d6 --- /dev/null +++ b/experiments/length-CPU360-final.csv @@ -0,0 +1,46 @@ +label,experiment,cols,rows,length,op,rep,rtt_us,settle_us,bytes +CPU360-final,length,0,0,0,insert,0,4654,16016,141 +CPU360-final,length,0,0,0,insert,1,4378,10489,80 +CPU360-final,length,0,0,0,insert,2,4332,10601,81 +CPU360-final,length,0,0,0,insert,3,4435,10512,81 +CPU360-final,length,0,0,0,insert,4,4366,10438,81 +CPU360-final,length,0,0,0,insert,5,4362,10559,81 +CPU360-final,length,0,0,0,insert,6,4320,10559,81 +CPU360-final,length,0,0,0,insert,7,4327,10506,81 +CPU360-final,length,0,0,0,insert,8,4329,10533,81 +CPU360-final,length,0,0,20,insert,0,4371,10658,81 +CPU360-final,length,0,0,20,insert,1,4435,10483,81 +CPU360-final,length,0,0,20,insert,2,4431,10521,81 +CPU360-final,length,0,0,20,insert,3,4407,17349,158 +CPU360-final,length,0,0,20,insert,4,4370,10529,80 +CPU360-final,length,0,0,20,insert,5,4407,10547,81 +CPU360-final,length,0,0,20,insert,6,4400,10558,81 +CPU360-final,length,0,0,20,insert,7,4425,10539,81 +CPU360-final,length,0,0,20,insert,8,4480,10644,81 +CPU360-final,length,0,0,40,insert,0,4452,10652,81 +CPU360-final,length,0,0,40,insert,1,4506,10507,81 +CPU360-final,length,0,0,40,insert,2,4445,10609,81 +CPU360-final,length,0,0,40,insert,3,4509,10769,81 +CPU360-final,length,0,0,40,insert,4,4455,10535,81 +CPU360-final,length,0,0,40,insert,5,4492,10556,81 +CPU360-final,length,0,0,40,insert,6,4479,17386,158 +CPU360-final,length,0,0,40,insert,7,4447,10470,80 +CPU360-final,length,0,0,40,insert,8,4458,10702,81 +CPU360-final,length,0,0,80,insert,0,4534,10584,81 +CPU360-final,length,0,0,80,insert,1,4505,10713,81 +CPU360-final,length,0,0,80,insert,2,4506,10615,81 +CPU360-final,length,0,0,80,insert,3,4507,10669,81 +CPU360-final,length,0,0,80,insert,4,4593,10818,81 +CPU360-final,length,0,0,80,insert,5,4578,10757,81 +CPU360-final,length,0,0,80,insert,6,4530,10699,81 +CPU360-final,length,0,0,80,insert,7,4555,10660,81 +CPU360-final,length,0,0,80,insert,8,4512,10691,81 +CPU360-final,length,0,0,160,insert,0,4716,10842,81 +CPU360-final,length,0,0,160,insert,1,4662,10811,81 +CPU360-final,length,0,0,160,insert,2,4665,10787,81 +CPU360-final,length,0,0,160,insert,3,4665,10771,81 +CPU360-final,length,0,0,160,insert,4,4691,10800,81 +CPU360-final,length,0,0,160,insert,5,4643,10841,81 +CPU360-final,length,0,0,160,insert,6,4653,10806,81 +CPU360-final,length,0,0,160,insert,7,4670,10899,81 +CPU360-final,length,0,0,160,insert,8,4690,10890,81 diff --git a/experiments/length-CPU360-fine.csv b/experiments/length-CPU360-fine.csv new file mode 100644 index 0000000..1cabf17 --- /dev/null +++ b/experiments/length-CPU360-fine.csv @@ -0,0 +1,46 @@ +label,experiment,cols,rows,length,op,rep,rtt_us,settle_us,bytes +CPU360-fine,length,0,0,0,insert,0,4684,15988,141 +CPU360-fine,length,0,0,0,insert,1,4438,10511,80 +CPU360-fine,length,0,0,0,insert,2,4375,10539,81 +CPU360-fine,length,0,0,0,insert,3,4441,10641,81 +CPU360-fine,length,0,0,0,insert,4,4417,10594,81 +CPU360-fine,length,0,0,0,insert,5,4396,10583,81 +CPU360-fine,length,0,0,0,insert,6,4340,10492,81 +CPU360-fine,length,0,0,0,insert,7,4367,10471,81 +CPU360-fine,length,0,0,0,insert,8,4391,10472,81 +CPU360-fine,length,0,0,20,insert,0,4401,10650,81 +CPU360-fine,length,0,0,20,insert,1,4419,10660,81 +CPU360-fine,length,0,0,20,insert,2,4427,10583,81 +CPU360-fine,length,0,0,20,insert,3,4496,17324,158 +CPU360-fine,length,0,0,20,insert,4,5204,10552,80 +CPU360-fine,length,0,0,20,insert,5,4391,12938,81 +CPU360-fine,length,0,0,20,insert,6,4401,10649,81 +CPU360-fine,length,0,0,20,insert,7,4442,10688,81 +CPU360-fine,length,0,0,20,insert,8,4387,10635,81 +CPU360-fine,length,0,0,40,insert,0,4405,10608,81 +CPU360-fine,length,0,0,40,insert,1,4404,10576,81 +CPU360-fine,length,0,0,40,insert,2,4443,10690,81 +CPU360-fine,length,0,0,40,insert,3,4417,10706,81 +CPU360-fine,length,0,0,40,insert,4,4416,10656,81 +CPU360-fine,length,0,0,40,insert,5,4541,10646,81 +CPU360-fine,length,0,0,40,insert,6,4497,17324,158 +CPU360-fine,length,0,0,40,insert,7,4471,10567,80 +CPU360-fine,length,0,0,40,insert,8,4493,10735,81 +CPU360-fine,length,0,0,80,insert,0,4578,10838,81 +CPU360-fine,length,0,0,80,insert,1,4605,10844,81 +CPU360-fine,length,0,0,80,insert,2,4524,10767,81 +CPU360-fine,length,0,0,80,insert,3,4585,10665,81 +CPU360-fine,length,0,0,80,insert,4,4666,10826,81 +CPU360-fine,length,0,0,80,insert,5,4608,10706,81 +CPU360-fine,length,0,0,80,insert,6,4564,10756,81 +CPU360-fine,length,0,0,80,insert,7,4586,10794,81 +CPU360-fine,length,0,0,80,insert,8,4530,10629,81 +CPU360-fine,length,0,0,160,insert,0,4618,10778,81 +CPU360-fine,length,0,0,160,insert,1,4708,10889,81 +CPU360-fine,length,0,0,160,insert,2,4715,10946,81 +CPU360-fine,length,0,0,160,insert,3,4712,10854,81 +CPU360-fine,length,0,0,160,insert,4,4653,10940,81 +CPU360-fine,length,0,0,160,insert,5,4655,10975,81 +CPU360-fine,length,0,0,160,insert,6,4639,10993,81 +CPU360-fine,length,0,0,160,insert,7,4684,10835,81 +CPU360-fine,length,0,0,160,insert,8,4691,10915,81 diff --git a/experiments/length-CPU360.csv b/experiments/length-CPU360.csv new file mode 100644 index 0000000..62f2f95 --- /dev/null +++ b/experiments/length-CPU360.csv @@ -0,0 +1,36 @@ +label,experiment,cols,rows,length,op,rep,rtt_us,settle_us,bytes +CPU360,length,0,0,0,insert,0,4640,16152,141 +CPU360,length,0,0,0,insert,1,4401,10465,80 +CPU360,length,0,0,0,insert,2,4458,10618,81 +CPU360,length,0,0,0,insert,3,4349,10504,81 +CPU360,length,0,0,0,insert,4,4445,10554,81 +CPU360,length,0,0,0,insert,5,4369,10652,81 +CPU360,length,0,0,0,insert,6,4397,10498,81 +CPU360,length,0,0,20,insert,0,4371,10644,81 +CPU360,length,0,0,20,insert,1,4393,10545,81 +CPU360,length,0,0,20,insert,2,4390,10548,81 +CPU360,length,0,0,20,insert,3,4378,10585,81 +CPU360,length,0,0,20,insert,4,4463,10590,81 +CPU360,length,0,0,20,insert,5,4465,17451,158 +CPU360,length,0,0,20,insert,6,4418,10533,80 +CPU360,length,0,0,40,insert,0,4506,10670,81 +CPU360,length,0,0,40,insert,1,4496,10680,81 +CPU360,length,0,0,40,insert,2,4452,10645,81 +CPU360,length,0,0,40,insert,3,4521,10660,81 +CPU360,length,0,0,40,insert,4,4421,10655,81 +CPU360,length,0,0,40,insert,5,4561,10801,81 +CPU360,length,0,0,40,insert,6,4471,10641,81 +CPU360,length,0,0,80,insert,0,4532,10709,81 +CPU360,length,0,0,80,insert,1,4566,10717,81 +CPU360,length,0,0,80,insert,2,4528,10678,81 +CPU360,length,0,0,80,insert,3,4553,10733,81 +CPU360,length,0,0,80,insert,4,4507,10805,81 +CPU360,length,0,0,80,insert,5,4547,10773,81 +CPU360,length,0,0,80,insert,6,4536,10637,81 +CPU360,length,0,0,160,insert,0,4720,10867,81 +CPU360,length,0,0,160,insert,1,4677,10832,81 +CPU360,length,0,0,160,insert,2,4673,10873,81 +CPU360,length,0,0,160,insert,3,4646,10752,81 +CPU360,length,0,0,160,insert,4,4729,17477,158 +CPU360,length,0,0,160,insert,5,4640,10674,80 +CPU360,length,0,0,160,insert,6,4679,10795,81 diff --git a/src/hal/clkrst.zig b/src/hal/clkrst.zig index 89a27ef..9902796 100644 --- a/src/hal/clkrst.zig +++ b/src/hal/clkrst.zig @@ -263,3 +263,93 @@ pub fn init(comptime p: Peripheral) void { setClockEnabled(p, true); resetPeripheral(p); } + +// --------------------------------------------------------------------------- the CPU's own clock + +/// Raise the HP CPU clock from the 90 MHz the bootloader leaves to `mhz`. +/// +/// WHY THIS IS CHEAP. The CPLL is ALREADY at 360 MHz: 90 is exactly 360/4, and the stock +/// second-stage bootloader gets there by setting `CONFIG_BOOTLOADER_CPU_CLK_FREQ_MHZ = 90` +/// (`bootloader_support/src/bootloader_clock_init.c:27-37`). So this is a divider change and +/// nothing else - no PLL to enable, no lock to wait for, and on the P4 no voltage step exists to +/// order it against (`esp_hw_support/port/esp32p4/rtc_clk_init.c:58-80` sets HP_ACTIVE DBIAS once +/// from efuse and never per-frequency). +/// +/// WHAT IT DOES NOT DISTURB, which is the reason it is safe to do from a running console: +/// * UART0's baud clock is selected by `PERI_CLK_CTRL110[25:24]` from XTAL, RC_FAST or PLL_F80M +/// (`hal/uart.zig:116-139`) - never the CPU clock. The console keeps its rate. +/// * The systimer is XTAL/2.5 = 16 MHz (`hal/systimer.zig:31`), so every timeout built on +/// `nowMs` keeps meaning what it meant. +/// * The flash interface runs from SPLL 480 MHz (`spimem_flash_ll.h:676-684`), so code executing +/// from flash-mapped memory is unaffected and this need not run from RAM. +/// * The `cycle` CSR counts real CPU cycles, so it simply counts faster. Nothing on the board +/// caches a cycles-per-microsecond figure; the HOST divisor in `experiments/` must move. +/// +/// The divider set and the ORDER are ESP-IDF's, from `rtc_clk_cpu_freq_to_cpll_mhz` +/// (`esp_hw_support/port/esp32p4/rtc_clk.c`). Only three CPU frequencies are legal on pre-v3 +/// silicon and each pins MEM/SYS/APB with it, because MEM must stay <= 200 MHz and APB <= 100: +/// +/// CPU 360 = CPLL/1, MEM = CPU/2 = 180, SYS = MEM/1 = 180, APB = SYS/2 = 90 +/// CPU 180 = CPLL/2, MEM = CPU/1 = 180, SYS = MEM/1 = 180, APB = SYS/2 = 90 +/// CPU 90 = CPLL/4, MEM = CPU/1 = 90, SYS = MEM/1 = 90, APB = SYS/1 = 90 +/// +/// APB lands at 90 MHz in all three, which is why peripherals do not care. Upscaling walks +/// APB -> SYS -> MEM -> CPU with a bus update after each: IDF's comment is explicit that the other +/// order passes through an intermediate state where APB or MEM violates its timing, and anything +/// touching those clocks during it may fault. +pub const CpuFreq = enum(u16) { mhz90 = 90, mhz180 = 180, mhz360 = 360 }; + +pub fn setCpuFreq(target: CpuFreq) void { + const root0 = Reg.at(regs.HP_SYS_CLKRST_ROOT_CLK_CTRL0_REG); + const root1 = Reg.at(regs.HP_SYS_CLKRST_ROOT_CLK_CTRL1_REG); + const root2 = Reg.at(regs.HP_SYS_CLKRST_ROOT_CLK_CTRL2_REG); + + const cpu_div = Field.of(regs.HP_SYS_CLKRST_REG_CPU_CLK_DIV_NUM_S, regs.HP_SYS_CLKRST_REG_CPU_CLK_DIV_NUM_V); + const cpu_num = Field.of(regs.HP_SYS_CLKRST_REG_CPU_CLK_DIV_NUMERATOR_S, regs.HP_SYS_CLKRST_REG_CPU_CLK_DIV_NUMERATOR_V); + const cpu_den = Field.of(regs.HP_SYS_CLKRST_REG_CPU_CLK_DIV_DENOMINATOR_S, regs.HP_SYS_CLKRST_REG_CPU_CLK_DIV_DENOMINATOR_V); + const mem_div = Field.of(regs.HP_SYS_CLKRST_REG_MEM_CLK_DIV_NUM_S, regs.HP_SYS_CLKRST_REG_MEM_CLK_DIV_NUM_V); + const sys_div = Field.of(regs.HP_SYS_CLKRST_REG_SYS_CLK_DIV_NUM_S, regs.HP_SYS_CLKRST_REG_SYS_CLK_DIV_NUM_V); + const apb_div = Field.of(regs.HP_SYS_CLKRST_REG_APB_CLK_DIV_NUM_S, regs.HP_SYS_CLKRST_REG_APB_CLK_DIV_NUM_V); + const update = Field.of(regs.HP_SYS_CLKRST_REG_SOC_CLK_DIV_UPDATE_S, regs.HP_SYS_CLKRST_REG_SOC_CLK_DIV_UPDATE_V); + + // Every divider register holds `divider - 1`. + const plan: struct { cpu: u32, mem: u32, sys: u32, apb: u32 } = switch (target) { + .mhz360 => .{ .cpu = 1, .mem = 2, .sys = 1, .apb = 2 }, + .mhz180 => .{ .cpu = 2, .mem = 1, .sys = 1, .apb = 2 }, + .mhz90 => .{ .cpu = 4, .mem = 1, .sys = 1, .apb = 1 }, + }; + + // The update bit is self-clearing and gates the whole divider set at once. Bounded, because an + // unbounded spin on a board with no debugger is indistinguishable from a crash. + const commit = struct { + fn go(r: Reg, f: Field) void { + r.modify(.{f.is(1)}); + _ = r.waitFor(f, 0, 100_000); + } + }.go; + + // Upscaling only: this firmware boots at 90 and never lowers. Doing it in the downscale order + // would leave APB above its 100 MHz limit while CPU was already fast. + root2.modify(.{apb_div.is(plan.apb - 1)}); + commit(root0, update); + root1.modify(.{sys_div.is(plan.sys - 1)}); + commit(root0, update); + root1.modify(.{mem_div.is(plan.mem - 1)}); + commit(root0, update); + root0.modify(.{ cpu_div.is(plan.cpu - 1), cpu_num.is(0), cpu_den.is(0) }); + commit(root0, update); + + // The source mux is NOT covered by the update bit and must move last; it is already CPLL here, + // so this is a no-op that documents the requirement rather than a step that changes anything. + // + // Then tell the mask ROM, because `ets_delay_us` and anything else built on `g_ticks_per_us` + // would otherwise delay by the wrong factor. `ets_update_cpu_frequency` is the recalibrator + // (`esp32p4.rom.ld:32`, 0x4fc00044). + ets_update_cpu_frequency(@intFromEnum(target)); +} + +/// Declared here rather than reached through `soc.rom`, because `soc` imports `hal` and the edge +/// cannot run both ways. It is a bare linker symbol either way - `build.zig` defines the address +/// once for the whole image - so a second declaration of it costs nothing and keeps the frequency +/// change and its recalibration in one function, where forgetting the second is impossible. +extern fn ets_update_cpu_frequency(mhz: u32) void; diff --git a/src/pardes/app.zig b/src/pardes/app.zig index 2f13865..5b1431f 100644 --- a/src/pardes/app.zig +++ b/src/pardes/app.zig @@ -185,6 +185,15 @@ export fn zig_main() noreturn { @as(u32, @intCast(heap.len / 1024)), }); + // The CPU clock, before anything is timed against it. The bootloader leaves 90 MHz and the + // CPLL is already at 360, so this is a divider change that disturbs neither UART0 (XTAL) nor + // the systimer (XTAL/2.5) nor the flash interface (SPLL). See hal/clkrst.zig:setCpuFreq. + if (config.cpu_mhz != 90) hal.clkrst.setCpuFreq(switch (config.cpu_mhz) { + 180 => .mhz180, + 360 => .mhz360, + else => .mhz90, + }); + const rwdt_was_armed = hal.rwdt.disable(); hal.systimer.init(); _ = rwdt_was_armed; @@ -208,6 +217,26 @@ export fn zig_main() noreturn { }); while (true) {} } + + // The CPU clock, measured rather than assumed. Every cycle count this firmware reports is + // divided by it somewhere, and `src/io/chip.zig` records it as "a measured ~90 MHz" that + // nothing here reconfigures - so it is worth printing rather than remembering. The systimer is + // XTAL/2.5 = 16 MHz and is NOT derived from the CPU clock (`hal/systimer.zig:31`, + // `clk_tree_defs.h:196-198`), which is exactly what makes it a valid reference for measuring it. + if (prof) { + const t_start = hal.systimer.micros(.unit0) orelse 0; + const c_start = soc.cycles(); + // 50 ms is long enough that the systimer's 16 MHz granularity and the loop's own overhead + // are both noise, and short enough to be invisible in a boot. + while ((hal.systimer.micros(.unit0) orelse 0) -% t_start < 50_000) {} + const elapsed_us = (hal.systimer.micros(.unit0) orelse 0) -% t_start; + const elapsed_cy = soc.cycles() - c_start; + soc.rom.print("MARK CPU_HZ cycles=%u us=%u khz=%u\r\n", .{ + @as(u32, @intCast(elapsed_cy)), + @as(u32, @intCast(elapsed_us)), + @as(u32, @intCast(if (elapsed_us > 0) elapsed_cy * 1000 / elapsed_us else 0)), + }); + } soc.rom.print("MARK PARDES_READY\r\n", .{}); var in: [256]u8 = undefined; diff --git a/src/soc.zig b/src/soc.zig index 350948a..1e94874 100644 --- a/src/soc.zig +++ b/src/soc.zig @@ -94,6 +94,11 @@ pub const rom = struct { pub extern fn ets_printf(fmt: [*:0]const u8, ...) c_int; pub extern fn ets_delay_us(us: u32) void; + /// Tell the ROM the CPU's new frequency, in MHz. `ets_delay_us` and everything else built on + /// `g_ticks_per_us` busy-waits by a cycle count derived from it, so a clock change without this + /// makes every ROM delay wrong by exactly the ratio. `esp32p4.rom.ld:32`, 0x4fc00044. + pub extern fn ets_update_cpu_frequency(mhz: u32) void; + /// Invalidate the caches. `map` selects which, from `rom/cache.h:228-236`: /// L1 ICache0 = 1, ICache1 = 2, L1 DCache = 0x10, L2 = 0x20; `cache_all` is all four. /// diff --git a/tools/rtt.zig b/tools/rtt.zig index 6ab0d87..d971c74 100644 --- a/tools/rtt.zig +++ b/tools/rtt.zig @@ -62,9 +62,14 @@ pub fn roundTrip( if (now - t0 > timeout_us) return null; } else if (now - last > quiet_us) break; - // Poll in millisecond units because that is what poll(2) takes; the TIMING above is - // microseconds and independent of this granularity. - const n = try port.readTimeout(&buf, 1); + // NON-BLOCKING, and that matters. `poll(2)` takes a timeout in MILLISECONDS, so waiting + // even 1 ms quantises this measurement into 1 ms buckets: a response that arrived just after + // a poll returned empty is reported up to a millisecond late. Against a round trip of a few + // milliseconds that is not a rounding error, it is a large fraction of the answer - measured + // as ~0.75 ms of unexplained gap between this figure and the sum of the board's own stage + // timings. A spin costs a busy host CPU for a few milliseconds per sample, which is the + // right trade for an instrument. + const n = try port.readTimeout(&buf, 0); if (n == 0) continue; if (first < 0) first = nowUs(port); bytes += n; |
