fix clang format workflow, update hackrf submodule, fix subghzd compilation warning, remove flashsize.h (#2992)

* trying to fix clang format workflow
* try without docker repo, classic apt install
* find what installed clang format 18 version is
* removed auto generated flashsize
* comment on flashsize.h
* modified to use repo owned clang-format binaries
* code format for windows hosts
* add clang-format binaries for linux x86_64/arm64, macos-arm64, windows x86_64
* add helper script to download specific llvm clang-format version
* add helper to download missing libraries (use docker)
* use custom clang-format
* fix ui_subghzd.cpp:85:29: warning: narrowing conversion of 'this->ui::SubGhzDRecentEntryDetailView::cnt' from 'uint32_t' {aka 'long unsigned int'} to 'uint16_t' {aka 'short unsigned int'} [-Wnarrowing]
* remove auto generated file
* Update tools/clang-format.sh
* removed fallback
* updated hackrf submodule

firmware/application/keeloq_file.cpp

@@ -43,7 +43,7 @@ bool read_keeloq_file(const fs::path& path, KeeloqData& data) {
 
     data.mf_name = std::string{chunks[0]};
     data.serial = std::strtoul(chunks[1].data(), NULL, 16);
-    data.counter = (uint16_t)std::atoi(chunks[2].data());
+    data.counter = std::atoi(chunks[2].data());
     data.btn = (uint8_t)std::atoi(chunks[3].data());
 
     return true;

firmware/application/keeloq_file.hpp

@@ -30,7 +30,7 @@
 struct KeeloqData {
     std::string mf_name{};
     uint32_t serial = 0;
-    uint16_t counter = 0;
+    uint32_t counter = 0;
     uint8_t btn = 0;
 };
 

firmware/flashsize.h

@@ -1,9 +0,0 @@
-#pragma once
-// DO NOT EDIT: IT IS AUTO GENERATED BY CMAKE!!!!
-
-// clang-format off
-//Allowed fw size in MB
-#define FLASH_SIZE_MB 4
-//Current compiled fw size in MB
-#define FLASH_SIZE_LIMIT_MB 3.5
-// clang-format on