fix(aspacem): raise amd64 managed address ceiling to 1 TiB

[not-matthias]

Raises amd64 aspacem_maxAddr from 128 GiB to 1 TiB so a client can float a 256 GiB MAP_NORESERVE reservation (e.g. snmalloc) under Callgrind instead of getting EINVAL. Scoped to VGP_amd64_linux.

Refs COD-3073

[greptile-apps]

Greptile Summary

Raises the managed address ceiling for VGP_amd64_linux from 128 GiB to 1 TiB in aspacemgr-linux.c, enabling clients that rely on large lazy (MAP_NORESERVE) reservations (e.g. snmalloc under Callgrind) to obtain a contiguous hole of ~512 GiB instead of only ~64 GiB.

• The change is correctly gated to #if defined(VGP_amd64_linux) inside the existing #if VG_WORDSIZE == 8 block; all other 64-bit Linux targets keep the 128 GiB value.
• The constant 0x10000000000ULL is 2^40 = 1 TiB, confirmed within amd64's 47-bit user-space VA; the ENABLE_INNER safety cap that trims aspacem_maxAddr to sp_at_startup still applies unchanged after the new block.
• Page-alignment assertions at startup (VG_IS_PAGE_ALIGNED(aspacem_maxAddr + 1)) continue to hold since 1 TiB is page-aligned.

Confidence Score: 5/5

Safe to merge; the change is a single, narrowly-scoped constant bump behind a platform guard with all existing invariants preserved.

The modification is one additional preprocessor branch setting a constant. The value (1 TiB = 2^40) fits well inside the 47-bit amd64 user-space VA, the ENABLE_INNER safety cap that clamps the ceiling to sp_at_startup still applies after the new block, and the page-alignment assertion at startup continues to pass because 0x10000000000 is page-aligned. All other 64-bit Linux targets are unaffected by the else branch.

No files require special attention; the single changed file is the right place for this constant.

Important Files Changed

Filename	Overview
coregrind/m_aspacemgr/aspacemgr-linux.c	Adds a VGP_amd64_linux-specific branch that sets aspacem_maxAddr to 1 TiB (0xFFFFFFFFFF) instead of 128 GiB; ENABLE_INNER cap and page-alignment invariants remain intact; hex constant and comments are correct.

Flowchart

%%{init: {'theme': 'neutral'}}%%
flowchart TD
    A[VG_am_startup called] --> B{VG_WORDSIZE == 8?}
    B -- No --> C[aspacem_maxAddr = sp_at_startup - 1]
    B -- Yes --> D{VGP_amd64_linux?}
    D -- Yes --> E[aspacem_maxAddr = 1 TiB - 1]
    D -- No --> F[aspacem_maxAddr = 128 GiB - 1]
    E --> G{ENABLE_INNER?}
    F --> G
    G -- Yes --> H[Cap at VG_PGROUNDDN sp_at_startup - 1]
    G -- No --> I[Keep value as-is]
    H --> J[aspacem_vStart = midpoint]
    I --> J
    J --> K[amd64_linux: vStart ~512 GiB / client hole ~512 GiB]

Loading

%%{init: {'theme': 'base', 'themeVariables': {"darkMode": true, "background": "#0d1117", "primaryColor": "#21262d", "primaryTextColor": "#e6edf3", "primaryBorderColor": "#8b949e", "lineColor": "#8b949e", "textColor": "#e6edf3", "edgeLabelBackground": "#161b22", "actorBkg": "#21262d", "actorBorder": "#8b949e", "actorTextColor": "#e6edf3", "actorLineColor": "#8b949e", "signalColor": "#8b949e", "signalTextColor": "#e6edf3", "noteBkgColor": "#373320", "noteBorderColor": "#d4a72c", "noteTextColor": "#f0e6c0", "labelBoxBkgColor": "#21262d", "labelBoxBorderColor": "#8b949e", "labelTextColor": "#e6edf3", "loopTextColor": "#e6edf3", "activationBkgColor": "#30363d", "activationBorderColor": "#8b949e"}}}%%
flowchart TD
    A[VG_am_startup called] --> B{VG_WORDSIZE == 8?}
    B -- No --> C[aspacem_maxAddr = sp_at_startup - 1]
    B -- Yes --> D{VGP_amd64_linux?}
    D -- Yes --> E[aspacem_maxAddr = 1 TiB - 1]
    D -- No --> F[aspacem_maxAddr = 128 GiB - 1]
    E --> G{ENABLE_INNER?}
    F --> G
    G -- Yes --> H[Cap at VG_PGROUNDDN sp_at_startup - 1]
    G -- No --> I[Keep value as-is]
    H --> J[aspacem_vStart = midpoint]
    I --> J
    J --> K[amd64_linux: vStart ~512 GiB / client hole ~512 GiB]

Loading

_{Reviews (1): Last reviewed commit: "fix(aspacem): raise amd64 managed addres..." | Re-trigger Greptile}

[codspeed-hq]

Merging this PR will not alter performance

✅ 42 untouched benchmarks
⏩ 88 skipped benchmarks¹

---

_{Comparing cod-3073-investigate-valgrind-support-for-snmallocs-256-gib (5714dad) with master (ce9d871)}

Footnotes

1. 88 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩