Key stability improvements made (deterministic + bounded)
1) Bounded memory usage in long-running stats
Fixed unbounded growth in NetworkState latency tracking:
Replaced std::vector<uint16_t> m_latency + push_back() with a fixed-size ring buffer (kLatencyWindow = 1024) and explicit counters.
Median latency computation now operates on at most 1024 samples, preventing memory growth and avoiding performance cliffs from ever-growing copies/sorts.
2) Prevent crash/UAF on shutdown + more predictable teardown
Controller shutdown ordering (Controller::stop()):
Now stops m_miner before destroying m_network.
This reduces chances of worker threads submitting results into a network listener that’s already destroyed.
Thread teardown hardening (backend/common/Thread.h):
Destructor now checks std::thread::joinable() before join().
Avoids std::terminate() if a thread object exists but never started due to early exit/error paths.
3) Fixed real leaks (including executable memory)
Executable memory leak fixed (crypto/cn/CnCtx.cpp):
CnCtx::create() allocates executable memory for generated_code via VirtualMemory::allocateExecutableMemory(0x4000, ...).
Previously CnCtx::release() only _mm_free()’d the struct, leaking the executable mapping.
Now CnCtx::release() frees generated_code before freeing the ctx.
GPU verification leak fixed (net/JobResults.cpp):
In getResults() (GPU result verification), a cryptonight_ctx was created via CnCtx::create() but never released.
Added CnCtx::release(ctx, 1).
4) JobResults: bounded queues + backpressure + safe shutdown semantics
The old JobResults could:
enqueue unlimited std::list items (m_results, m_bundles) → unbounded RAM,
call uv_queue_work per async batch → unbounded libuv threadpool backlog,
delete handler directly while worker threads might still submit → potential crash/UAF.
Changes made:
Hard queue limits:
kMaxQueuedResults = 4096
kMaxQueuedBundles = 256
Excess is dropped (bounded behavior under load).
Async coalescing:
Only one pending async notification at a time (m_pendingAsync), reducing eventfd/uv wake storms.
Bounded libuv work scheduling:
Only one uv_queue_work is scheduled at a time (m_workScheduled), preventing CPU starvation and unpredictable backlog.
Safe shutdown:
JobResults::stop() now detaches global handler first, then calls handler->stop().
Shutdown detaches m_listener, clears queues, and defers deletion until in-flight work is done.
Defensive bound on GPU result count:
Clamp count to 0xFF inside JobResults as well, not just in the caller, to guard against corrupted kernels/drivers.
5) Idempotent cleanup
VirtualMemory::destroy() now sets pool = nullptr after delete:
prevents accidental double-delete on repeated teardown paths.
Verification performed
codespell . --config ./.codespellrc: clean
CMake configure + build completed successfully (Release build)
Signed-off-by: rezky_nightky <with.rezky@gmail.com>
armv8l is what CMAKE_SYSTEM_PROCESSOR is set to when an ARMv8 processor
is in 32-bit mode, so it should be added to the ARMv7 target list even
though it's v8 because it's 32 bits. Currently, it's not in any ARM
target list which means x86 is assumed and the build fails.
Previously with HTTP API enabled on brenchmarking run, it is possible
to cause a segfault due to an issue handling the m_httpd pointer and
rebinding.
- Initialize m_httpd to nullptr to indicate when it's not in use.
- Safely delete m_httpd in Api's destructor to prevent use-after-free
issues.
- Add checks to ensure m_httpd is not nullptr before usage in start,
stop, and tick methods.
- Log errors for HTTP server start failures to aid in debugging.
FixesMoneroOcean/xmrig#120
Signed-off-by: Dave Walker (Daviey) <email@daviey.com>
If the caller knows the difficulty of a PoW hash a given nonce
yields, it can tell the callee via the X-Hash-Difficulty, which
may allow the callee to skip some processing if the difficulty
does not meet some criterion.
In my case, a merge mining proxy can know it's pointless trying
to submit the nonce to a chain with higher difficulty when the
nonce only meets the difficulty for a lower difficulty chain.
- Added average hashrate display
- Fixed the number of threads shown at startup
- Fixed `--threads` or `-t` command line option (but `--cpu-max-threads-hint` is recommended to use)
2021-11-27 12:27:26 +01:00
483 changed files with 42322 additions and 34593 deletions
- [#3346](https://github.com/xmrig/xmrig/pull/3346) ARM64 JIT: don't use `x18` register.
- [#3348](https://github.com/xmrig/xmrig/pull/3348) Update to latest `sse2neon.h`.
- [#3356](https://github.com/xmrig/xmrig/pull/3356) Updated pricing record size for **Zephyr** solo mining.
- [#3358](https://github.com/xmrig/xmrig/pull/3358) **Zephyr** solo mining: handle multiple outputs.
# v6.20.0
- Added new ARM CPU names.
- [#2394](https://github.com/xmrig/xmrig/pull/2394) Added new CMake options `ARM_V8` and `ARM_V7`.
- [#2830](https://github.com/xmrig/xmrig/pull/2830) Added API rebind polling.
- [#2927](https://github.com/xmrig/xmrig/pull/2927) Fixed compatibility with hwloc 1.11.x.
- [#3060](https://github.com/xmrig/xmrig/pull/3060) Added x86 to `README.md`.
- [#3236](https://github.com/xmrig/xmrig/pull/3236) Fixed: receive CUDA loader error on Linux too.
- [#3290](https://github.com/xmrig/xmrig/pull/3290) Added [Zephyr](https://www.zephyrprotocol.com/) coin support for solo mining.
# v6.19.3
- [#3245](https://github.com/xmrig/xmrig/issues/3245) Improved algorithm negotiation for donation rounds by sending extra information about current mining job.
- [#3254](https://github.com/xmrig/xmrig/pull/3254) Tweaked auto-tuning for Intel CPUs.
- [#3271](https://github.com/xmrig/xmrig/pull/3271) RandomX: optimized program generation.
XMRig is a high performance, open source, cross platform RandomX, KawPow, CryptoNight and AstroBWT unified CPU/GPU miner and [RandomX benchmark](https://xmrig.com/benchmark). Official binaries are available for Windows, Linux, macOS and FreeBSD.
XMRig is a high performance, open source, cross platform RandomX, KawPow, CryptoNight and [GhostRider](https://github.com/xmrig/xmrig/tree/master/src/crypto/ghostrider#readme) unified CPU/GPU miner and [RandomX benchmark](https://xmrig.com/benchmark). Official binaries are available for Windows, Linux, macOS and FreeBSD.
## Mining backends
- **CPU** (x64/ARMv8)
- **CPU** (x86/x64/ARMv7/ARMv8/RISC-V)
- **OpenCL** for AMD GPUs.
- **CUDA** for NVIDIA GPUs via external [CUDA plugin](https://github.com/xmrig/xmrig-cuda).
If you want use HTTP API you need enable it (`"enabled": true,`) then choice `port` and optionaly `host`. API not available if miner built without HTTP support (`-DWITH_HTTP=OFF`).
If you want use HTTP API you need enable it (`"enabled": true,`) then choice `port` and optionally `host`. API not available if miner built without HTTP support (`-DWITH_HTTP=OFF`).
Offical HTTP client for API: http://workers.xmrig.info/
Official HTTP client for API: http://workers.xmrig.info/
@@ -17,7 +17,7 @@ Double check that you see `Huge pages 100%` both for dataset and for all threads
### Benchmark with custom config
You can run benchmark with any configuration you want. Just start without command line parameteres, use regular config.json and add `"benchmark":"1M",` on the next line after pool url.
You can run benchmark with any configuration you want. Just start without command line parameters, use regular config.json and add `"benchmark":"1M",` on the next line after pool url.
# Stress test
@@ -26,4 +26,4 @@ You can also run continuous stress-test that is as close to the real RandomX min
xmrig --stress
xmrig --stress -a rx/wow
```
This will require Internet connection and will run indefinitely.
This will require Internet connection and will run indefinitely.
- [#268](https://github.com/xmrig/xmrig-amd/pull/268) [#270](https://github.com/xmrig/xmrig-amd/pull/270) [#271](https://github.com/xmrig/xmrig-amd/pull/271) [#273](https://github.com/xmrig/xmrig-amd/pull/273) [#274](https://github.com/xmrig/xmrig-amd/pull/274) [#1171](https://github.com/xmrig/xmrig/pull/1171) Added RandomX support for OpenCL, thanks [@SChernykh](https://github.com/SChernykh).
- Algorithm `cn/wow` removed, as no longer alive.
- Algorithm `cn/wow` removed, as no longer alive.
# v3.2.0
- Added per pool option `coin` with single possible value `monero` for pools without algorithm negotiation, for upcoming Monero fork.
@@ -103,7 +103,7 @@
- [#1105](https://github.com/xmrig/xmrig/issues/1105) Improved auto configuration for `cn-pico` algorithm.
- Added commands `pause` and `resume` via JSON RPC 2.0 API (`POST /json_rpc`).
- Added command line option `--export-topology` for export hwloc topology to a XML file.
- Breaked backward compatibility with previous configs and command line, `variant` option replaced to `algo`, global option `algo` removed, all CPU related settings moved to `cpu` object.
- Broken backward compatibility with previous configs and command line, `variant` option replaced to `algo`, global option `algo` removed, all CPU related settings moved to `cpu` object.
- Options `av`, `safe` and `max-cpu-usage` removed.
- Algorithm `cn/msr` renamed to `cn/fast`.
- Algorithm `cn/xtl` removed.
@@ -122,7 +122,7 @@
- [#1092](https://github.com/xmrig/xmrig/issues/1092) Fixed crash if wrong CPU affinity used.
- [#1103](https://github.com/xmrig/xmrig/issues/1103) Improved auto configuration for RandomX for CPUs where L2 cache is limiting factor.
- [#1105](https://github.com/xmrig/xmrig/issues/1105) Improved auto configuration for `cn-pico` algorithm.
- [#1106](https://github.com/xmrig/xmrig/issues/1106) Fixed `hugepages` field in summary API.
- [#1106](https://github.com/xmrig/xmrig/issues/1106) Fixed `hugepages` field in summary API.
- Added alternative short format for CPU threads.
- Changed format for CPU threads with intensity above 1.
- Name for reference RandomX configuration changed to `rx/test` to avoid potential conflicts in future.
@@ -150,7 +150,7 @@
- [#1050](https://github.com/xmrig/xmrig/pull/1050) Added RandomXL algorithm for [Loki](https://loki.network/), algorithm name used by miner is `randomx/loki` or `rx/loki`.
- Added [flexible](https://github.com/xmrig/xmrig/blob/evo/doc/CPU.md) multi algorithm configuration.
- Added unlimited switching between incompatible algorithms, all mining options can be changed in runtime.
- Breaked backward compatibility with previous configs and command line, `variant` option replaced to `algo`, global option `algo` removed, all CPU related settings moved to `cpu` object.
- Broken backward compatibility with previous configs and command line, `variant` option replaced to `algo`, global option `algo` removed, all CPU related settings moved to `cpu` object.
- Options `av`, `safe` and `max-cpu-usage` removed.
- Algorithm `cn/msr` renamed to `cn/fast`.
- Algorithm `cn/xtl` removed.
@@ -183,7 +183,7 @@
- [#314](https://github.com/xmrig/xmrig-proxy/issues/314) Added donate over proxy feature.
- Added new option `donate-over-proxy`.
- Added real graceful exit.
# v2.14.4
- [#992](https://github.com/xmrig/xmrig/pull/992) Fixed compilation with Clang 3.5.
- [#1012](https://github.com/xmrig/xmrig/pull/1012) Fixed compilation with Clang 9.0.
@@ -250,16 +250,16 @@
# v2.8.1
- [#768](https://github.com/xmrig/xmrig/issues/768) Fixed build with Visual Studio 2015.
- [#769](https://github.com/xmrig/xmrig/issues/769) Fixed regression, some ANSI escape sequences was in log with disabled colors.
- [#777](https://github.com/xmrig/xmrig/issues/777) Better report about pool connection issues.
- [#777](https://github.com/xmrig/xmrig/issues/777) Better report about pool connection issues.
- Simplified checks for ASM auto detection, only AES support necessary.
- Added missing options to `--help` output.
# v2.8.0
- **[#753](https://github.com/xmrig/xmrig/issues/753) Added new algorithm [CryptoNight variant 2](https://github.com/xmrig/xmrig/issues/753) for Monero fork, thanks [@SChernykh](https://github.com/SChernykh).**
- Added global and per thread option `"asm"` and and command line equivalent.
- Added global and per thread option `"asm"` and command line equivalent.
- **[#758](https://github.com/xmrig/xmrig/issues/758) Added SSL/TLS support for secure connections to pools.**
- Added per pool options `"tls"` and `"tls-fingerprint"` and command line equivalents.
- [#767](https://github.com/xmrig/xmrig/issues/767) Added config autosave feature, same with GPU miners.
- [#767](https://github.com/xmrig/xmrig/issues/767) Added config autosave feature, same with GPU miners.
- [#245](https://github.com/xmrig/xmrig-proxy/issues/245) Fixed API ID collision when run multiple miners on same machine.
- [#153](https://github.com/xmrig/xmrig/issues/153) Fixed issues with dwarfpool.com.
# v2.4.1
- [#147](https://github.com/xmrig/xmrig/issues/147) Fixed comparability with monero-stratum.
@@ -371,7 +371,7 @@
- [#101](https://github.com/xmrig/xmrig/issues/101) Fixed MSVC 2017 (15.3) compile time version detection.
- [#108](https://github.com/xmrig/xmrig/issues/108) Silently ignore invalid values for `donate-level` option.
- [#111](https://github.com/xmrig/xmrig/issues/111) Fixed build without AEON support.
# v2.3.1
- [#68](https://github.com/xmrig/xmrig/issues/68) Fixed compatibility with Docker containers, was nothing print on console.
@@ -398,7 +398,7 @@
# v2.1.0
- [#40](https://github.com/xmrig/xmrig/issues/40)
Improved miner shutdown, fixed crash on exit for Linux and OS X.
- Fixed, login request was contain malformed JSON if username or password has some special characters for example `\`.
- Fixed, login request was contain malformed JSON if username or password has some special characters for example `\`.
- [#220](https://github.com/fireice-uk/xmr-stak-cpu/pull/220) Better support for Round Robin DNS, IP address now always chosen randomly instead of stuck on first one.
- Changed donation address, new [xmrig-proxy](https://github.com/xmrig/xmrig-proxy) is coming soon.
@@ -418,16 +418,16 @@ Improved miner shutdown, fixed crash on exit for Linux and OS X.
- Fixed Windows XP support.
- Fixed regression, option `--no-color` was not fully disable colored output.
- Show resolved pool IP address in miner output.
# v1.0.1
- Fix broken software AES implementation, app has crashed if CPU not support AES-NI, only version 1.0.0 affected.
# v1.0.0
- Miner complete rewritten in C++ with libuv.
- This version should be fully compatible (except config file) with previos versions, many new nice features will come in next versions.
- This is still beta. If you found regression, stability or perfomance issues or have an idea for new feature please fell free to open new [issue](https://github.com/xmrig/xmrig/issues/new).
- This version should be fully compatible (except config file) with previous versions, many new nice features will come in next versions.
- This is still beta. If you found regression, stability or performance issues or have an idea for new feature please fell free to open new [issue](https://github.com/xmrig/xmrig/issues/new).
- Added new option `--print-time=N`, print hashrate report every N seconds.
- New hashrate reports, by default every 60 secons.
- New hashrate reports, by default every 60 seconds.
- Added Microsoft Visual C++ 2015 and 2017 support.
- Removed dependency on libcurl.
- To compile this version from source please switch to [dev](https://github.com/xmrig/xmrig/tree/dev) branch.
@@ -440,7 +440,7 @@ Improved miner shutdown, fixed crash on exit for Linux and OS X.
- Fixed gcc 7.1 support.
# v0.8.1
- Added nicehash support, detects automaticaly by pool URL, for example `cryptonight.eu.nicehash.com:3355` or manually via option `--nicehash`.
- Added nicehash support, detects automatically by pool URL, for example `cryptonight.eu.nicehash.com:3355` or manually via option `--nicehash`.
# v0.8.0
- Added double hash mode, also known as lower power mode. `--av=2` and `--av=4`.
@@ -124,7 +124,7 @@ Force enable (`true`) or disable (`false`) hardware AES support. Default value `
Mining threads priority, value from `1` (lowest priority) to `5` (highest possible priority). Default value `null` means miner don't change threads priority at all. Setting priority higher than 2 can make your PC unresponsive.
#### `memory-pool` (since v4.3.0)
Use continuous, persistent memory block for mining threads, useful for preserve huge pages allocation while algorithm switching. Possible values `false` (feature disabled, by default) or `true` or specific count of 2 MB huge pages. It helps to avoid loosing huge pages for scratchpads when RandomX dataset is updated and mining threads restart after a 2-3 days of mining.
Use continuous, persistent memory block for mining threads, useful for preserve huge pages allocation while algorithm switching. Possible values `false` (feature disabled, by default) or `true` or specific count of 2 MB huge pages. It helps to avoid losing huge pages for scratchpads when RandomX dataset is updated and mining threads restart after a 2-3 days of mining.
#### `yield` (since v5.1.1)
Prefer system better system response/stability `true` (default value) or maximum hashrate `false`.
@@ -133,7 +133,7 @@ Prefer system better system response/stability `true` (default value) or maximum
Enable/configure or disable ASM optimizations. Possible values: `true`, `false`, `"intel"`, `"ryzen"`, `"bulldozer"`.
#### `argon2-impl` (since v3.1.0)
Allow override automatically detected Argon2 implementation, this option added mostly for debug purposes, default value `null` means autodetect. This is used in RandomX dataset initialization and also in some other mining algorithms. Other possible values: `"x86_64"`, `"SSE2"`, `"SSSE3"`, `"XOP"`, `"AVX2"`, `"AVX-512F"`. Manual selection has no safe guards - if your CPU doesn't support required instuctions, miner will crash.
Allow override automatically detected Argon2 implementation, this option added mostly for debug purposes, default value `null` means autodetect. This is used in RandomX dataset initialization and also in some other mining algorithms. Other possible values: `"x86_64"`, `"SSE2"`, `"SSSE3"`, `"XOP"`, `"AVX2"`, `"AVX-512F"`. Manual selection has no safe guards - if your CPU doesn't support required instructions, miner will crash.
#### `astrobwt-max-size`
AstroBWT algorithm: skip hashes with large stage 2 size, default: `550`, min: `400`, max: `1200`. Optimal value depends on your CPU/GPU
xmrig
SChernykh
Willie Shen
rezky_nightky
HashVault
Tony Wang
user0-07161
slayingripper
Christopher Wright
Ben Westover
ybh1998
Ikko Eltociear Ashimine
Matthew Vogel
goodmost
Dave Walker (Daviey)
moneromooo-monero
SChernykh
Jackson Zheng
Tony Butler
Jeremy Chadwick
Dmitriy Nikiforov
Matthew Smith
Bartłomiej Burdukiewicz
benthetechguy