Skip to content

Uminer GPU v0.1.25 (Testing Release)

Latest
Latest

Choose a tag to compare

View all tags
@github-actions github-actions released this 16 Jul 02:58
v0.1.25
732ef9c

Uminer GPU v0.1.25 Testing Release

English

Highlights

  • Adds always-on local search-domain protection for long-lived jobs, repeated notifications,
    reconnects and pool A-to-B-to-A recovery, reducing accidental repeated GPU work without requiring
    a private pool extension.
  • Hardens third-party Stratum compatibility by keeping optional hashrate reporting endpoint-scoped,
    session-scoped and fail-open. Unsupported reporting methods do not stop mining.
  • Strengthens BTX/CSD shared-session behavior, bounded protocol output and per-session accounting
    while preserving one user-pool socket for multi-GPU rigs.

Minimal Usage

# BTX
./uminer --coin btx -o stratum+tcp://global.btxpool.org:23333 -u <BTX_ADDRESS>.rig01 -p x

# CSD; tcp/tls is selected by the URL and is never forced globally.
./uminer --coin csd -o stratum+tls://csd.luckypool.io:443 -u <CSD_ADDRESS>.rig01 -p x

Changes

  • Work cursors are owned by the effective work domain. A known domain resumes from its prior
    cursor rather than restarting after a reconnect or repeated job, and an exhausted domain is not
    silently rewound.
  • The protection is automatic and cannot be disabled by a user option. Standard third-party pools
    remain compatible; a pool-specific work-domain field is treated only as an optional validated
    enhancement.
  • Periodic hashrate reporting state is isolated by session, endpoint and worker identity. The
    default auto mode does not blindly send private methods to an unknown pool.
  • Explicit reporting profiles remain available for pools that document support. Method-not-found
    or a live-session timeout disables only that optional reporting dialect; GPU work and share
    submission continue.
  • CSD supports negotiated extranonce2 widths and exhaustively advances the extranonce2/nonce
    search space without silent truncation. BTX and CSD continue to use the shared bounded runtime.
  • Pool-supplied numeric JSON-RPC IDs may use a number or a canonical decimal string where the
    selected profile permits it; unknown IDs remain harmless and do not alter A/S/R accounting.

Compatibility

  • Linux x86_64 only; Windows is not included.
  • glibc 2.30 or newer.
  • Embedded native SASS: sm_75, sm_80, sm_86, sm_89, sm_120.
  • NVIDIA driver 535 or newer for Turing/CMP, Ampere and Ada.
  • NVIDIA driver 580.95 or newer for Blackwell.
  • No CUDA Toolkit installation is required at runtime.
  • Compatibility remains pool-profile specific. This release does not claim compatibility with
    every Stratum implementation solely from fake-pool or transport tests.

Developer Fee

  • BTX: 2%.
  • CSD: 3%.
  • The per-coin policy is compiled into this release and cannot be reduced through the CLI.
  • A GPU changes to developer work only after a trusted Fee route has connected, authorized and
    supplied a valid job. If no route reaches READY, user mining continues and the missed window is
    not recovered. The user-pool connection remains online during a Fee window.
  • The configured direct Fee fallback routes include unencrypted TCP endpoints because those pool
    endpoints do not provide TLS. Plain TCP does not provide transport confidentiality or endpoint
    authentication.

Verification and Scope

The frozen private-source candidate passed 248 workspace tests with no failures, strict lint,
architecture, license and RustSec gates. Two independent clean Active-Fee CUDA release builds were
byte-identical in pre-release qualification. This remains a Linux testing-stage release under an
explicit fast-release exception; ongoing real-pool, GPU-matrix and long-duration production
qualification is not claimed here.

The local guard prevents a single Uminer process from rewinding known work. Absolute uniqueness
across every independent miner requires a pool-assigned global namespace and is not claimed for
third-party pools that do not provide one.

繁體中文

重點

  • 新增始終啟用的本機搜尋域保護，涵蓋長時間未換區塊、重複通知、重連及礦池 A→B→A
    回復，無需私有礦池擴充即可降低 GPU 重複計算。
  • 強化第三方 Stratum 相容性：可選算力自報按端點、會話及身份隔離，且採 fail-open；礦池不支援
    自報方法時不會停止挖礦。
  • 強化 BTX/CSD 共用會話、有限協議輸出與逐會話記帳，同時維持多卡整機只有一條使用者礦池
    socket。

最簡使用方式

# BTX
./uminer --coin btx -o stratum+tcp://global.btxpool.org:23333 -u <BTX_ADDRESS>.rig01 -p x

# CSD；tcp/tls 完全由 URL 選擇，不做全域強制。
./uminer --coin csd -o stratum+tls://csd.luckypool.io:443 -u <CSD_ADDRESS>.rig01 -p x

變更

  • 工作游標歸屬於實際工作域。已知工作域在重連或重複任務後會從先前游標繼續，不會重新起跑；
    已耗盡的工作域也不會被靜默回捲。
  • 此保護自動啟用且使用者無法關閉。標準第三方礦池保持相容；礦池專用 work-domain 欄位只作為
    可選且經驗證的增強能力。
  • 週期算力自報狀態按會話、端點與 worker 身份隔離。預設 auto 不會向未知礦池盲目送出私有
    方法。
  • 對已明確記錄支援能力的礦池，仍可選擇指定自報 profile。收到 method-not-found 或活會話逾時
    只會停用該可選方言；GPU 工作及 share 提交持續進行。
  • CSD 支援協商後的 extranonce2 寬度，並完整推進 extranonce2/nonce 搜尋空間，不做靜默截斷；
    BTX 與 CSD 繼續共用同一有限 runtime。
  • 當所選 profile 允許時，礦池回傳的 JSON-RPC 數字 ID 可使用數值或規範十進位字串；未知 ID
    不會污染 A/S/R 記帳。

相容性

  • 僅提供 Linux x86_64，不含 Windows。
  • glibc 2.30 或更新版本。
  • 內嵌 native SASS：sm_75、sm_80、sm_86、sm_89、sm_120。
  • Turing/CMP、Ampere、Ada 需 NVIDIA 535 或更新驅動。
  • Blackwell 需 NVIDIA 580.95 或更新驅動。
  • 執行時不需安裝 CUDA Toolkit。
  • 相容性仍以礦池 profile 為準；本版不會僅憑 fake-pool 或 transport 測試宣稱相容所有
    Stratum 實作。

開發者費用

  • BTX：2%。
  • CSD：3%。
  • 各幣種費率已固化於本版本，CLI 無法下調。
  • 只有受信任 Fee 路由完成連線、授權並取得有效任務後，GPU 才會切換。若沒有路由進入
    READY，使用者挖礦持續且該時間窗不追補；Fee 時間窗內使用者礦池連線保持在線。
  • 已設定的 Fee 直接備援路由包含未加密 TCP 端點，因為這些礦池端點未提供 TLS。明文 TCP
    不提供傳輸機密性或端點身份驗證。

驗證與發佈邊界

凍結的私有源碼候選已通過 248 個 workspace 測試且零失敗，並通過嚴格 lint、架構、授權及
RustSec 門禁。發佈前資格檢查中的兩次獨立 clean Active-Fee CUDA release build 位元組完全
一致。本版仍依明確快速發佈特例提供 Linux 測試版；持續進行中的真實礦池、GPU 矩陣與長時間
生產資格驗收不在本版宣稱範圍內。

本機保護可避免單一 Uminer 進程回捲已知工作；若第三方礦池未提供全域命名空間，則不宣稱
所有獨立礦機之間具有絕對唯一性。

Assets 10
Loading