Activity since 2025-06-29

• esp-rs/esp-hal repo issues and pull requests

  • Security: non-random stack smashing canary severely restricts its effectiveness #3667 [gibbz00] bug

    Currently, __stack_chk_guard seems to be "hardcoded" with via the ESP_HAL_CONFIG_STACK_GUARD_VALUE. ...

    • gibbz00: The stack guard is mostly aimed at developers right now, not security. Ah ok, fair. Was going...

  • Cleanup esp-metadata #3686 [bugadani] skip-changelog

    This PR splits out the driver-specific configs and tries to explain them a bit.

    • MabezDev: Thanks!

  • Improved stack overflow protection on RISC-V targets #3674 [gibbz00]

    esp-hal has supports stack overflow protection using flip-link on esp32c6 and esp32h2, (unsure why it's limited to...

    • gibbz00: Glad to hear 😊 I even think it would be possible to get a hacky/patchy way for stack probing...

    • bjoernQ: Glad to hear 😊 I even think it would be possible to get a hacky/patchy way for stack probing...

  • Add toolchain arg to more xtask commands #3714 [bugadani]

    Why? Because I screwed up the previous change, that relied on RUSTUP_TOOLCHAIN. Obviously that env var always...

    • JurajSadel

    • bjoernQ: Passing in a toolchain is much nicer than relying on whatever built the xtask

  • xtask: msrv-bump #3708 [bjoernQ]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • JurajSadel

    • bugadani

  • Automate MSRV bumps #3611 [MabezDev] automation

    We're going to do this often, and it's actually listed in a lot of places. We should automate this step. Things to...

  • Update `xtensa-toolchain` #3719 [SergioGasquez]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bjoernQ: LGTM

    • JurajSadel

  • fix: increase esp32 DRAM by 16K #3703 [gibbz00]

    This is mostly an experimentation right now to check if CI passes with these changes, hence the draft. Think it's ok...

    • bjoernQ: upping the available memory is probably fine - not sure if this is now really easier to read or not

    • gibbz00: Nice :) Chose 0x3FFAE000 for it to be easier to cross-reference values in the TRM. Not sure if...

    • bugadani

    • bjoernQ: I mean "ORIGIN = 0x3FFAE000 + 8K + RESERVE_DRAM, len = 192K - RESERVE_DRAM" looks "confusing" in...

    • gibbz00: Ah ok. Agreed. There's also this one...

  • test #3721 [bugadani] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

  • Minor linker script touchups. #3717 [gibbz00] skip-changelog

    Inlines rodata_desc.x and consolidates .trap output section definitions. Changes aren't necessarily impactful....

    • bjoernQ: Pretty much all cosmetic surgery - not against doing it

    • bugadani

  • chore(esp-hal/ld): include metadata.x with the other shared sections #3673 [gibbz00] skip-changelog

    Don't think this needs a changelog entry.

    • bjoernQ: Probably a matter of taste - honestly, I don't have an opinion on this. Just LGTMing this, so...

  • Simplify peripheral resets #3720 [bugadani] skip-changelog

    This PR updates PACs to get rid of some unnecessary special casing. The PR also replaces the set/clear duplication...

    • bugadani: Better to assert before underflowing ref_count. Although maybe this should be a debug_assert?

      • bjoernQ: Not sure - maybe important enough for a general assert (and given this is not a hot function)

    • bjoernQ: Nice improvement

    • MabezDev: Thank you for this, this has been bothering me for a while, and caused me pain when playing...

  • Tweak and make use of the old `bind_X_interrupt` functions #3689 [bugadani] skip-changelog

    This PR changes the naming scheme of bind_X_interrupts so that we can now define a common function name for common...

    • JurajSadel

    • MabezDev: Thanks!

  • Place I2C async interrupt handler into ram, wait for bus_busy to actually clear #3722 [bugadani]

    • bugadani: I guess we'll see if this (as in, waiting for bus_busy) is actually a fix, or if I've just...

    • MabezDev: I wonder if we should be doing this for all our async handlers? 🤔 They are generally very small,...

      • bugadani: Yeah this isn't bad, and seeing just how much esp-idf forces into IRAM we could be a bit more...

    • MabezDev: Thanks!

  • Move UART instance info to esp-metadata #3698 [bugadani] skip-changelog

    • copilot-pull-request-reviewer: Pull Request Overview This PR moves UART instance metadata into the esp-metadata crate, adds a...

      • copilot-pull-request-reviewer: The function name generate_uart_peripehrals contains a typo (peripehrals); consider renaming it...

      • copilot-pull-request-reviewer: The function name generate_uart_peripehrals contains a typo; rename to generate_uart_peripherals...

    • MabezDev: Thanks!

  • fix: increase esp32 heap by about 15K #3705 [gibbz00]

    This is mostly an experimentation right now to check if CI passes with these changes, hence the draft. Size increase...

    • bjoernQ: I don't think this is fine - we need to leave the app-core stack alone according to ...

    • gibbz00: Ahh. So when processor core runs whilst app core is starting, app ROM startup stack may still...

    • bjoernQ: Yes that's my understanding - after the second core is started we could reclaim the memory but...

  • Fix defmt timestamp #3725 [bugadani]

    fn _esp_println_timestamp is documented as expected to return millis, but the defmt timestamp was printed as micros.

    • bjoernQ

  • Adjust ESP32-S2 and ESP32-S3 memory region lengths to reflect those defined in ESP-IDF #3709 [gibbz00]

    The size reduction in #734 based on #732 (comment) seems to unfortunately be outdated. The change in...

    • bjoernQ: Reducing DRAM/IRAM seg is probably good - the next version of espflash will bring in updated...

    • gibbz00: Reducing DRAM/IRAM seg is probably good - the next version of espflash will bring in updated...

    • bjoernQ: Oki. A similar change would also need to be applied to ESPS3's memory.x no? From the commit in...

    • gibbz00: From the commit in ESP-IDF you linked to - yes - thanks for noticing. Ok! Added this in this...

  • Extract SPI metadata #3702 [bugadani]

    Additionally, this PR bundles the SIO signals into two slices similar to CS, macro-generates their respective...

    • copilot-pull-request-reviewer: Pull Request Overview This PR extracts and improves SPI metadata generation while bundling SIO...

      • copilot-pull-request-reviewer: Typo in function name: 'generate_spi_slave_peripehrals' should be...

      • copilot-pull-request-reviewer: Typo in function name: 'generate_spi_master_peripehrals' should be...

    • MabezDev: Thanks, very strange about the semver checks error 🤔

  • Fix mcpwm docs #3727 [bugadani] documentation skip-changelog

    H2 can only output <= XTAL frequency, we shouldn't display a higher value in the docs, using that would result in an error.

    • MabezDev: Thanks!

  • Handling of different bootloaders / bootloader versions #3327 [bjoernQ] RFC

    Current Situation the HAL won't include an app-descriptor which is needed by later versions of the esp-idf 2nd...

    • MabezDev: I'm pretty confident that we can resolve this post 1.0, it mostly relies on figuring out a way...

  • Calibration calc overflow #3090 [yuri-ncs]

    Submission Checklist 📝 I have updated existing examples or added new ones (if applicable). I have used cargo...

    • MabezDev: Thanks for taking a look at this! Closing for now, but please open a new PR if you plan on...

  • Resolve `C-EXAMPLE`: Uart #3712 [SergioGasquez] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: I don't think we need to explain into_blocking/into_async like this. It's trivial once the...

      • MabezDev: Fair enough, in which case we should write this documentation on Blocking and Async and then...

      • bjoernQ: Fair enough, in which case we should write this documentation on Blocking and Async and then...

      • SergioGasquez: I created general examples for into_blocking/async and linked to them from here, let me know if...

    • bugadani: Should we maybe add a single "basic usage" example instead of this?

      • MabezDev: We have a basic usage example on the Uart struct itself, and on Uart::new. I don't see these...

    • MabezDev

      • MabezDev: Let's remove the spaces here, it makes the generated docs look a bit odd

        • MabezDev: Oh I just realised you commented out the init code in these, imo, based on my comment #3712...

      • MabezDev: I don't think we need to call with_tx when documenting with_rx, and vice versa.

        • MabezDev: same with cts/rts example too, we only need rx + cts, and then tx + rts

      • MabezDev: Let's inline the Config::default() chain into apply_config, no need to separate this out into a variable.

      • MabezDev: Nit: let's just call these let uart = , and maybe we should use peripherals.UART0 instead of UART1?

    • MabezDev: I forget, does no_run still get tested with our doc tests?

      • bugadani: it'll build

    • bugadani: You can put a single # to stop rendering the row, but still place some visual separation into...

    • MabezDev: LGTM with one change, thank you!

      • MabezDev: I think part can just reference the Blocking struct again. Something like "see Blocking for more details"

  • Introduce esp-rom-sys crate #3688 [bjoernQ] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • MabezDev

      • MabezDev: This feels a bit flaky. We're passing linker directives via a dummy config?

        • MabezDev: Is it? cfg.insert inserts this into the config, won't this show up in the generated config table...

        • bjoernQ: ah yes - I thought of the way how we passed it from esp-rom-sys to esp-hal Yes, here we used "a...

      • MabezDev: Can't we just put this in the linker script in esp-hal by default?

        • MabezDev: We can always move this around behind the scenes later, if we get a nice solution for adding to...

      • MabezDev: Could you update the xtask so that this crate never gets a release beyond 0.1? Only 0.1.x versions.

        • MabezDev: The problem with that is that our release workflow will fail, and then we manually have to...

        • bjoernQ: we probably should do both since we could also (for whatever reason) skip the xtask and change...

    • MabezDev: This is getting close, thanks for sticking with it! A few more comments from my side, then I...

      • MabezDev: This isn't the only place we call update_package, I suggest we do the check in there instead.

      • MabezDev: Imo we should just take the bump amount and and unconditionally change it to Patch when package...

        • bjoernQ: Can we just set the current version to 0.1.0 here now? I wasn't sure if the tooling needs to see...

      • MabezDev: I'd probably remove most of this, it's a bit waffly, the first couple of sentences summerize the...

    • MabezDev: I wouldn't return an error here, just log it and force the version to be a patch bump.

      • MabezDev: To be clear, I would ditch is_version_bump_acceptable and just do a match on the package, if...

      • bjoernQ: yeah - doesn't feel too good to have a "hack" based on a specific package in a "random"...

      • MabezDev: There would be a warning/info message, plus this gets manual review anyways when we open the...

      • bjoernQ: Ah I see your concerns. cargo xrelease plan will create the plan with the inappropriate...

      • MabezDev: Thanks, I see the workflow now. That should work. here is still one catch: if we start with...

      • bjoernQ: I'm still not convinced it's the best solution but implemented the suggestion to get this going...

      • bjoernQ: Wow that was a "race condition" .... it probably doesn't make much of a difference and one day...

    • MabezDev: Did you mean? Suggested change ...

      • bjoernQ: 🤦‍♂️ ofc ... it's 33C outside and it feels like 45C in my office so let's blame this on...

      • MabezDev: Heh, It's not just me suffering then :D, I'm boiling here 😫 . Thanks!

    • MabezDev: LGTM! Thanks for working on this, and getting it past the finish line. Once this is merged,...

    • bjoernQ: Thanks! Yes will do that after the merge - good idea to just manually release a 0.1.0 and...

  • ESP32: psram & esp-storage conflict #3642 [mjwells2002] bug

    Bug description writing to onboard flash corrupts values stored in psram To Reproduce write to onboard flash while...

  • de-`paste!` the `any_peripheral` macro #3699 [bugadani] skip-changelog

    • MabezDev: Took me a second to wrap my head around this :D. So now all Any related bits belong in...

  • Use esp-idf rom-function patches in esp-storage for esp32 #3644 [bjoernQ] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bjoernQ: Closing in favor of #3688

  • Add sdm #2371 [katyo]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • MabezDev: GPIO, and the hal has changed a lot. I would love to see this added, but I think a new PR is...

  • Fix broken link #3729 [bugadani] documentation skip-changelog

    • MabezDev: Thanks!

  • Refactor AssistDebug symbols #3730 [bugadani] skip-changelog

    • MabezDev: LGTM, thanks!

  • esp-config TUI #3649 [bjoernQ] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: I still think this is bad UX and we shouldn't let users select conflicting choices in the first place.

      • MabezDev: I think this is resolved now? In my testing, on esp-wifi's rx_queue, entering letters or a...

      • bugadani: The comment can be resolved by removing the loop or by adding a comment why it's still needed ;)

      • bjoernQ: I kept the loop "just in case" - removed

    • bugadani: I'd like to avoid listing every chip here. Can we iterate through features, try to parse the...

    • bugadani: This is still missing the octal format somehow.

    • bugadani: Suggested change ...

    • bugadani: This would be less confusing as -> Result<(), String>

    • bugadani: Take Option<&str>. We shouldn't be this liberal with cloning/allocating memory for no good reason.

    • bugadani: Suggested change work_dir:...

    • MabezDev: .or_else(chip_from_config), does this work?

    • MabezDev: Thanks for working on this! My concerns have been resolved, LGTM!

  • Prepare changelog #3731 [bjoernQ] skip-changelog

    Prepare the esp-rom-sys CHANGELOG after the in-between release see #3688 (review)

    • bugadani

    • SergioGasquez: We should probably add it to the list in...

  • Move `rom` module from esp-hal to new crate, esp-rom #3389 [MabezDev]

    Following on from #3354 (comment) and something I've had on my mind for a little while too. ROM code is...

  • Light-/deep-sleep support for ESP32-H2 #1029 [jessebraham] low power status:long-term chip:esp32h2

    This will serve as a tracking issue for light-/deep-sleep support for the ESP32-H2. Please feel free to use this...

    • philipcristiano: Ah awesome, is there a way to sponsor / bounty the issue?

  • Generate peripherals macro call & remove paste #3713 [bugadani] skip-changelog

    This PR removes the hand-maintained peripherals! macro call and generates it from metadata instead. Peripheral...

    • bugadani: There are a few places where symbols are named using the PAC peripheral name (RTC_CNTL), and...

    • bugadani: This is not great but we'll want to distance the LP IOMUX functionalities a bit anyway from HP...

    • bugadani: Semver-check is mad becaue this PR marks the GPIO singleton stable - which is totally fair. The...

    • MabezDev: -1330, nice :D. LGTM, thanks for all the effort on this quest line, getting this far is a nice milestone!

  • RFC: Use `libesp_rom.a` for ESP32 in esp-storage #3529 [bjoernQ] RFC package:esp-storage

    In esp-storage we cannot just use the ROM functions like we do for all other chips - we need "patched" versions of...

    • bjoernQ: Closed via #3688

  • `esp-println` has no output #3732 [KortanZ] bug status:needs-attention

    Bug description using esp-generate generate a very simple project, add esp-println, then using println! in code. No...

    • bjoernQ: Which development board are you using and how did you connect it? (UART bridge vs JTAG-serial)

    • KortanZ: Waveshare ESP32-C6-DEV-KIT-N8 development board, it has both UART birdge and jtag. I'm using UART bridge.

    • bjoernQ: The board apparently comes with an onboard USB-hub which might confuse the auto-detection in...

    • KortanZ: The board apparently comes with an onboard USB-hub which might confuse the auto-detection in...

  • Improve esp-wifi cfg docs #3733 [bjoernQ] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • MabezDev: Thanks!

    • bugadani: I'm not sure when we started generating a scrollbar, but this looks ... not great.

  • Validate wifi configuration #1611 [bugadani] package:esp-wifi

    esp-idf does some validation on the configuration, e.g. #if (CONFIG_ESP_WIFI_RX_BA_WIN >...

  • Tweak config table #3734 [bugadani] skip-changelog

    Let's see how this works out. Stability column is a bit meh, but the goal is to make this table somewhat readable,...

    • MabezDev: This looks a lot cleaner, thanks!

  • Remove magic #3735 [bugadani]

    • MabezDev: I think this might be something we want actually, the magic never sat quite right with me. At...

  • Fix and extend pin layouts #3726 [bugadani] skip-changelog

    This PR fixes various swapped U0RX/TX functions, adds missing functions based on TRMs, removes inaccessible pins...

    • MabezDev: Thanks!

  • Check for global symbols in esp-hal library #3738 [MabezDev] CI status:blocked automation

    Dependent on #3706. This issue is about ensuring we don't pollute the library name space with more global symbols....

  • RMT: Refactor state checks, support non-blocking poll on blocking transactions #3716 [wisp3rwind]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • MabezDev: Maybe this should return Event, and then we can add helper methods to determine whether it's...

      • wisp3rwind: Hmm, I'm not sure I follow you; what would be the advantage of returning an Event here? Would it...

      • wisp3rwind: Or is what you're proposing simply something along the lines of pub struct TransactionState { ...

      • bugadani: Do we need any of this in the first place? We generally don't support non-blocking poll in other...

      • wisp3rwind: One more note: I don't think it's all that useful for .poll() to convey whether the hardware is...

      • wisp3rwind: Do we need any of this in the first place? We generally don't support non-blocking poll in other...

      • bugadani: How does that work when the API itself doesn't seem to allow pushing data to an ongoing...

      • wisp3rwind: Currently, you do indeed need to provide all the data in the initial call that starts the...

      • bugadani: It also places the burden of chunking the data on user code Why? The driver can do the same...

      • wisp3rwind: It also places the burden of chunking the data on user code Why? The driver can do the same...

      • bugadani: What should the path forward for this PR be? Just to be clear, I'm not challenging anything you...

      • wisp3rwind: What should the path forward for this PR be? Just to be clear, I'm not challenging anything you...

      • Dominaezzz: The question of passing in data in .transmit() or .poll() is actually related to a thing I've...

  • RMT: Extend tx and rx capabilities and API #3509 [wisp3rwind]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

  • Initialize `syscall_table` in esp-rom-sys #3740 [bjoernQ]

    Remaining ideas from #3389 We should also consider properly initializing...

  • Lot's of different debug assertions with `esp-hal-1.0.0-beta.1` #3718 [0x53A] bug status:needs-attention

    Bug description I've tried to port a small project from IDF to no_std and am getting debug assertions. Different...

    • bugadani: My project is here Your reproducer seems to be private

    • 0x53A: apologies! should be public now

    • bugadani: You're reserving too much memory for heap which ends up causing memory corruption. Reserve less...

    • 0x53A: oops, well, thank you for taking a look! Do you have a pointer for me where I could read up on...

  • Post release rollover #3616 [MabezDev] skip-changelog

  • Resolve `C-EXAMPLE`: Spi #3736 [SergioGasquez] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: The formatting is a bit off here, we should format examples like rustfmt would format them.

      • SergioGasquez: Is there any way to run rustfmt for doc examples?

      • bugadani: https://rust-lang.github.io/rustfmt/?version=v1.8.0&search=format_code_in_doc_comments#format_cod...

      • SergioGasquez: In an application this would...

      • bugadani: Rustfmt should take doc_comment_code_block_width into account, it shouldn't break 73 characters...

      • SergioGasquez: https://github.com/esp-rs/esp-hal/actions/runs/16027751187/job/45220000749?pr=3736, maybe xtask...

  • Make cargo xfmt respect the format config #3739 [bugadani] skip-changelog

    This PR fixes #3736 (comment)

    • SergioGasquez: Thanks for looking into this! I just tried it with #3736 and still formats it in a weird way: ...

    • bugadani: and still formats it in a weird way: For me code in doc comments aren't touched at all. The...

    • SergioGasquez: LGTM!

  • Allow hooking into the executor #3737 [bugadani]

    This PR should enable both #3728 and #3670 without actually integrating them into the executor.

    • bjoernQ: I guess for s.th. like #3728 embassy-executor's rtos-trace feature (or embassy-rs/embassy#4033)...

    • bugadani: Yeah the only advantage of this approach is that this is free for executors that don't opt into...

    • bjoernQ: Code looks good - just wondering if it's worth to add a test for this

    • bugadani: What would you like to be tested?

    • bjoernQ: What would you like to be tested? Nothing fancy - just something exercising run_with_callbacks...

    • bjoernQ

  • fix(esp-hal): avoid disabling the same interrupt twice in `_setup_interrupts` #3741 [gibbz00]

    disable doesn't use the Cpu argument passed in: #[cfg(multi_core)] disable(Cpu::AppCpu, intr); disable(Cpu::ProCpu,...

    • bugadani: This will be incorrect when we start supporting the P4, until then there are no multi_core...

  • Resolve `C-EXAMPLE`: Spi #3743 [SergioGasquez] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: LOL I'm sorry

    • MabezDev: Thanks!

  • fix nightly ci #3662 [JurajSadel] skip-changelog

    closes #3659

  • Requesting feedback: Improve async wifi performance #3670 [karlri]

    This change seems to increase async wifi performance quite nicely. It helps with http response times and with ping...

    • bugadani: With #3737 merged, it should now be possible to implement this in user code. Maybe we should...

  • Add embassy executor cpu stats #3728 [Szybet]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: Suggested change esp-hal ...

      • Szybet: Without it it doesn't compile standalone. If that is intentional I will disable it

      • bugadani: You need to enable unstable in your own project, not like this.

      • Szybet: done

    • Szybet: As #3737 is merged do you wish to redo this PR as an example using hooking into executor? I...

    • bugadani: Sure, why not? rtos-trace is a bit better tool to gather usage stats in general, but this would...

  • Config options for specific chips #3020 [MabezDev] package:esp-config 1.0-blocker

    Following on from the discussion here: #3001 (comment) We might want a better solution than just documenting which...

    • bjoernQ: I explored the idea of individual config.toml files here:...

    • MabezDev: I explored the idea of individual config.toml files here:...

  • esp-storage may be unsound #3009 [MabezDev] bug package:esp-storage

    Relevant info here: #1902

    • MabezDev: Closed via #3677

  • Use 1.88 & bump HAL MSRV #3742 [bugadani] skip-changelog

    • bugadani: Skip-changelog added because of the clippy fix in esp-backtrace.

    • MabezDev: Thanks!

    • bugadani: Come on CI, stop failing

  • make peripheral module private #3745 [MabezDev] skip-changelog

    It no longer has public items since we no longer have any trait magic. So let's make it private. Technically a...

    • bugadani

  • Add macro to let rustfmt format code blocks in doc comments #3744 [bugadani] skip-changelog

    This PR adds a new proc macro that allows replacements in code blocks in doc comments. The macro is meant to allow...

    • bjoernQ: A shame something like this is needed. IIRC I also considered a proc-macro but "Cannot be an...

    • bugadani: Cannot be an inner attribute, so not applicable on module-level docs Well if we accept that...

    • bjoernQ: I like this approach - also nice to have custom replacements IIRC we always needed nightly for...

    • bjoernQ: Oh wait ... seems rustfmt is tripping over the syntax for me and doesn't format the code snippets 🙃

    • bugadani: Yeah looks like we need a space for rustfmt's sake, i.e. # {before_snippet}. Significant...

    • bugadani: This isn't ideal, but it renders as expected. Replacing in the middle of a line would be...

  • esp32s3 wifi station mode can scan but found nothing after wakeup from sleep #3751 [vult0306] bug low power package:esp-wifi

    Case 1: Start esp32s3 wifi dhcp example, can scan and connect to wifi normally. Case 2: When device is up, put...

    • 0xkelvin: after waking up from sleep, actually

    • MabezDev: Could you provide us with a minimal repro using the latest crate releases?

  • esp-config tui - multi config selection #3752 [MabezDev] package:esp-config

    Context in this thread: #3020

  • Resolve `C-EXAMPLE` violations #2623 [bugadani] documentation 1.0-blocker

    As documentation isn't as immutable as the API is, we may not treat rustdoc examples as strictly as the Rust API...

    • SergioGasquez: Should we also document AnyX clone_unchecked and reborrow methods? We could do something like we...

    • MabezDev: Should we also document AnyX clone_unchecked and reborrow methods? We could do something like we...

  • Rework how peripheral singletons are generated #3747 [bugadani] skip-changelog

    This PR remove the remaining paste use from the peripherals macros, and introduces a for_each macro for peripheral...

    • bugadani: This isn't required, there's just code relying on crate::pac to exist.

    • bugadani: Drive-by warning fix

    • bugadani: I'm not sure if we still need this. Interrupts now have their own functions in the peripheral...

    • bugadani: This is still not great, I actually need a macro that exposes the list of peripherals. The...

    • bugadani: This part (optionally generating fields and attributes) is probably way to much dark wizardry at...

      • Dominaezzz: I spent way too long staring at this but I'm convinced it works now. In essence, this is being...

    • Dominaezzz: Did you forget to delete this file?

      • bugadani: yup, thanks

    • Dominaezzz

    • bjoernQ: LGTM

  • Add app desc macro examples #3750 [SergioGasquez] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: Can you add a comment to each line what they are? The last 4 are rather hard to understand...

    • bjoernQ: maybe adding "Format: major * 100 + minor" here and for the maximum might be useful

    • bjoernQ: Thanks

    • bjoernQ: I guess in this case it's fine to use ignore instead of no_run - getting this to pass...

    • MabezDev: Thanks!

  • High-speed LEDC channels do not output signal on ESP32 #3753 [Axel-ex] bug peripheral:ledc chip:esp32

    Bug description High-speed LEDC channels configured via esp-hal on ESP32 do not generate any PWM signal, even though...

  • (re-)introduce the rt feature on esp-hal #3706 [MabezDev]

    cc: #2589 Adds the rt feature which disables esp_hal::init and removes runtime symbols from the global name space....

    • MabezDev: I think we should remove this, and put it back in esp-wifi, I don't like the idea of exporting...

      • bjoernQ: we can make sure we do what esp-idf does: storing the xtal freq in one of the RTC retention...

  • QOL: Do something about `regi2c_*` functions. #1740 [playfulFence] status:long-term

    It would be nice to do something about them regi2c_* functions (see example below), because honestly, every time I...

    • bugadani: Oh, I didn't know about this issue, is #3255 enough to close this?

    • playfulFence: I also somehow forgot to revisit the issue after that PR. Yeah, I think these changes make our...

  • Bluetooth client api/example #1606 [ccruquan] examples package:esp-wifi

    Is any plan support device as bluetooth client?

  • How does sleep work on multicore chips? #2108 [dimpolo] bug low power

    I noticed that calling Rtc::sleep_deep from the second core does not put the first core to sleep. Is this expected...

  • Sigma-delta modulation driver #2370 [katyo] peripheral:gpio

    Motivations Almost all ESP SoCs has Sigma-delta modulation periphery which can be used as a DAC...

  • Unexpected behaviour of `ram(rtc_fast, persistent)` #2516 [jukrut] bug low power chip:esp32

    Bug description I cloned the esp-hal repo and ran the ram.rs example with a little adjustment: - loop { - ...

  • I2C: Every line of code is a liability. Take some time to see if your implementation can be simplified before opening a PR. #2769 [bugadani] peripheral:i2c 1.0 non-breaking

    The hardware versions are seemingly more similar than the implementation suggests, we may want to explore this.

    • MabezDev: Whilst I agree with the sentiment, there isn't anything specifically actionable here. Closing.

  • Stable driver: I2C master #2493 [MabezDev] peripheral:i2c api

    As part of #2491, which has more details on driver analysis. esp-hal API-GUIDELINE omissions #2763 #2764 #2765 ...

  • Resolve `C-EXAMPLE`: Gpio and I2c #3746 [SergioGasquez] skip-changelog

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: I'd keep the imports visible, it's part of the use

    • bjoernQ: Thanks

  • documentation website has not been updated for 1.0.0-beta.0 #3309 [Be-ing] bug

    Bug description The documentation at https://docs.esp-rs.org/esp-hal/ for esp-hal still shows the documentation for...

    • bjoernQ: I guess I resolved this issue - try these links which should now redirect to reasonable...

    • bjoernQ: @MabezDev do you consider this done? Anything missing?

    • MabezDev: I think that's everything, thank you!

  • I2C should expose Ack settings for each transfer (like esp-idf does) #2505 [Dominaezzz] peripheral:i2c

    Motivations The I2C hardware can be used to speak SCCB which is a protocol similar to I2C used to speak to...

  • Allow splitting control over modem clocks. #3687 [Frostie314159]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: Do you think we can also remove the RADIO_CLK peripheral?

      • Frostie314159: I've removed it with the latest commit.

    • bugadani: Just a note for future improvement, this refcounter is now unnecessary, I think - we just need...

      • Frostie314159: I've reduced it to one.

    • bugadani: I think I'd prefer checked_sub and a panic (not a simple - to save binary size when using...

      • Frostie314159: I added an expect with an error message describing the problem.

    • bugadani: Can you please avoid auto-formatting TOML files? It's not great :D

      • Frostie314159: Oh crap, yeah, I'll fix it. I haven't yet figured out how to disable the extension for neovim.

    • bugadani: I'm pretty happy with how this looks, although I still have a few technicalities

      • bugadani: Let use the unwrap! macro here. You can still print a message, but it'll be defmt-formatted if...

      • bugadani: This function is not unsafe, so you don't need a # Safety comment. On the other hand, if this...

        • Frostie314159: I can add a method for the PhyClockGuard, which forgets it.

        • bugadani: please don't, let's not encourage bad behaviour

      • bugadani: These comments no longer needs the parens, as peripherals are now always passed by value

    • Frostie314159: What should I do about the migration guide?

    • bugadani

      • bugadani: Technically this isn't true anymore, this just track how many users keep the PHY clock alive :)

      • bugadani: This is fine as a fact, although I'm not sure how much value this comment has in the source :)

      • bugadani: This comment also seems outdated

      • bugadani: Please mark all new public api with instability::unstable. Could this maybe be called...

      • bugadani: Suggested change /// Unsafely...

      • bugadani: Suggested change /// If the ref...

      • bugadani: Suggested change /// Reset the...

    • bjoernQ: Kind of nitpick but all the comments like stealing X is safe since it is passed into init are...

    • bjoernQ: I see why but it looks a bit fishy to me: It doesn't matter which of the peripherals we steal...

      • Frostie314159: I could also export the internal functions, but with #[doc(hidden)].

      • bjoernQ: I guess it's fine code wise. We should just have a comment why we use potentially unrelated...

  • Provide a more robust solution for getting the current time #1944 [Dominaezzz]

    The current implementations for getting the time (since boot?)...

    • MabezDev: I think this will be addressed if we ever support another bootloader. We'll find out pretty fast...

  • ESP32 / ESP_WIFI: Problems when enabling COEX #2624 [bjoernQ] bug package:esp-wifi chip:esp32

    Originally mentioned on Matrix:...

    • MabezDev: Would be good to try this again soon.

  • More Examples #2995 [Makuo12]

    I just recently used embassy with esp-hal crate to work on a project that uses the matrix keypad and an lcd module....

  • Make `asynch` modules private #2319 [MabezDev] documentation

    Some drivers have this module public, some don't. Imo now async is enabled by default, we should make these modules...

    • MabezDev: Only I2S has one now, and its far from stabilization. We'll likely remove this when we start...

  • BLE and Wifi examples not working on ESP32S3. #1626 [hydra] package:esp-wifi chip:esp32s3

    Hi, I'm testing the examples on board here, and none of the wifi or bluetooth examples that I tried work. I was...

  • Better define (or document) the startup state of the cores #3410 [bjoernQ] documentation

    See #3403 (comment)

  • Separate `esp-riscv-rt` from `esp-hal` #3638 [gibbz00] linker scripts package:esp-riscv-rt

    Hey again, As discussed in the matrix chat 1, I'm investigating the possibility to run my ESP32C3 without needing to...

  • Introduce `esp-metadata-generated` #3754 [bugadani] skip-changelog

    This PR extracts generated code to a new crate. Long-term plans are that this crate is generated/updated from an...

    • bugadani: To be reverted, formatter decided to ignore comments

    • bugadani: Description: A `macro_rules!` declarative macro cannot be invoked by its prior name. The macro...

  • Fix nightly CI #3755 [JurajSadel] skip-changelog

    closes #3748

    • bugadani: This doesn't look equivalent to me. The left hand side went from 0..63 to 0..=1, the right hand...

      • JurajSadel: Good catch

    • JurajSadel: nightly CI run: https://github.com/JurajSadel/esp-hal/actions/runs/16078495660

  • dma: add SDIO dedicated DMA types #3749 [rmsyn]

    Adds types needed for SoCs with a dedicated SDIO DMA engine (e.g. ESP32-C6). Thank you for your contribution! We...

    • rmsyn: This is still a static mut, and I plan to create a new-type to wrap it. Probably something...

      • bugadani: Just FYI these macros are a bad idea (they are unsound as you can use safe code to get 2 mutable...

      • rmsyn: Right, I wasn't sure how important the macros were to the surrounding machinery in the dma...

    • bugadani

      • bugadani: Is the ESP32 substantially different?

        • rmsyn: It has a different bitfield layout for the DMA descriptor flags. I'll need to review the ESP32...

      • bugadani: If this is C6-specific, these don't apply.

      • bugadani: You should keep your notes in a tracking issue or the relevant PR's discussion.

        • rmsyn: 😅 yeah... that got pulled in on a stray git add esp-hal. I'll remove it.

      • bugadani: What's the idea here? I don't think we want the ability to share DMA descriptors with multiple...

        • rmsyn: My thinking was that the ESP32-C6 has two cores, which can each access the static DMA...

    • Dominaezzz: I took a look at the TRM. The SDIO's DMA is different enough to warrant it going into...

      • rmsyn: Sounds good, I can move this code into the main SDIO PR.

    • rmsyn: Closing this PR in favor of the suggestions in #3749 (comment) Moving the code into #3503

  • Unexpected breakage when migrating to 1.0.0-beta.0 #3252 [dimpolo] bug

    When trying to update esp-hal to 1.0.0-beta.0 a couple of required changes were difficult to discover because they...

    • dimpolo: "completed" as in there's nothing to do here or just that there's a new beta available? I just...

  • sdio: initial SDIO HAL implementation #3503 [rmsyn]

    Thank you for your contribution! We appreciate the time and effort you've put into this pull request. To help us...

    • bugadani: Why do you need to create pin drivers?

      • bugadani: let us know if you can find this in esp-idf, otherwise we'll ask around internally

      • rmsyn: After reviewing the available InputSignal/OutputSignal definitions in esp-metadata, there don't...

      • Dominaezzz: There aren't signals for the SDIO, it's IO_MUX only, like the USB_OTG signals. Have a look at...

      • rmsyn: I ended up finding the signals in the ESP32 config in esp-metadata. The config just leaves out...

      • bugadani: OH

    • bugadani: These should follow the naming in the TRM, i.e. SDIO_*. I'm adding these signals among others in...

      • rmsyn: For sure, I just followed the naming convention in the ESP32 config. I'll switch them out for...

      • rmsyn: I rebased on your changes, and updated the names in ESP32 to be consistent with ESP32C6. This...

    • rmsyn: I'm starting to implement the read_command/write_response, and the read/write_data functions....

    • bugadani: What do you all think? Sounds good as this one is already quite large To be fair you could...

    • rmsyn: To be fair you could put it here, too, purely additive PRs are much easier to deal with even if...

  • ESP-NOW never sends if Rtc::new is called afterwards #3756 [furrycoding] bug status:needs-attention

    Bug description If both ESP-NOW and low-power management (RTC) is used in a project, and...

created at 2025-07-06 02:19 by dinghy 0.13.1.