Termux: sendme 0.37.1 aborts with uninitialized Android context (release panic=abort)
Maintainers usually reply within 1 day
Nobody has claimed this yet.
Assessment
- Difficulty
- 4/5
- Estimated time
- 3-5 days
- Newbie friendliness
- 48/100
- Issue type
- Bug
- Clarity
- Mostly clear
- Activity status
- Active
- Tech stack
- android, rust
- Domain
- cli, networking
Research direction
Start by comparing the release profile in Cargo.toml with the Android context handling documented for iroh 1.3.0 and the cached ndk-context 0.1.1 source. Check how the resolver is initialized for standalone Termux use and whether a release build with unwinding avoids the SIGABRT; the report says that rebuild has not been tested. Done means both send and receive start successfully on Termux, including with relay disabled.
Written by the indexing model from the issue text.
Description
Sendme 0.37.1 immediately aborts in Termux on Android 14 when starting either send or receive. No ticket is emitted by the sender and no transfer completes. Android's crash log reports android context was not initialized (SIGABRT). This also happens with --relay disabled.
This may be a remaining variant of #133, which was closed with a request to retry 0.36.1 or later. Here the process exits completely with status 134 rather than just logging a recoverable thread panic. The published release profile uses panic = "abort", while iroh's Android DNS fallback documentation explicitly requires unwinding when no JNI context exists. Could the release profile or Android resolver initialization be adjusted for standalone Termux use?
AI-assisted reproduction, environment, and debugging evidence
These diagnostics were collected with AI assistance and reviewed before posting. Only newly generated synthetic files were used; tickets, ephemeral endpoint IDs, and private addresses are omitted.
Reproduce
printf 'sendme synthetic Termux smoke 2026-11-09\n' > phone-test.txt
sendme --version
# sendme 0.37.1
RUST_BACKTRACE=full RUST_LOG=trace sendme send --no-progress phone-test.txt
# process terminates immediately, shell exit 134 / Python returncode -6
RUST_BACKTRACE=full RUST_LOG=trace sendme send --no-progress --relay disabled phone-test.txt
# same SIGABRT
Receiving was also tested with a real ticket produced by sendme 0.37.1 on Linux x86_64 for a separate 42-byte synthetic file:
RUST_BACKTRACE=full RUST_LOG=trace sendme receive --no-progress '<fresh Linux sender ticket>'
# same immediate SIGABRT; receive directory remains empty
The Linux sender successfully imported its file and printed a ticket. The Android sender never reaches that stage, so Android-to-Linux download and end-to-end checksum verification are blocked. Default relay and disabled relay fail identically before file transfer. No permissions, security settings, or resolver configuration were changed.
Environment and provenance
- Google Pixel 4a (5G), Android 14 / API 34, arm64-v8a; Linux 4.19.278, aarch64.
- Termux 0.118.3 (F-Droid); termux-tools 1.45.0.
sendme 0.37.1, installed in$HOME/.cargo/bin/sendmethrough Cargo, registry sourceregistry+https://github.com/rust-lang/crates.io-index.- Cargo's
.crates2.jsonrecordsall_features=true,no_default_features=false,profile=release,target=aarch64-linux-android. The exact historical install command / whether--lockedwas used is not recorded. - Binary SHA256:
8e133c4dc4df994ef87b6db1d313f4fb6e20bcec0c10e1033025862bcd928585. - ELF aarch64 PIE, dynamically linked, interpreter
/system/bin/linker64, Android 24 minimum, NDK r29 (14206865), stripped. cargo 1.95.0 (f2d3ce0bd 2026-03-21),rustc 1.95.0 (59807616e 2026-04-14), hostaarch64-linux-android, LLVM 21.1.8; Termux source-built toolchain.LD_PRELOAD=$PREFIX/lib/libtermux-exec-ld-preload.so;LD_LIBRARY_PATHempty.- Test date: 2026-11-09. Latest GitHub release at test time: v0.37.1 (2026-10-21).
Dependency / source evidence
The installed crate's cached, packaged Cargo.lock SHA256 is 8aa0077b519a6bed8237bdb0b9d6fe1b4fa20fc970fcdbc668fd56fcae0d51a7. It contains iroh/iroh-base/iroh-dns/iroh-relay 1.3.0, iroh-blobs 0.103.1, ndk-context 0.1.1, and netwatch 0.19.3. Those sources are present locally. This is the packaged lockfile, not a preserved build receipt; exact historical resolution beyond Cargo's install metadata is unverified.
The packaged Cargo.toml has:
[profile.release]
opt-level = "s"
lto = true
codegen-units = 1
debug = 0
panic = "abort"
strip = true
The cached ndk-context-0.1.1/src/lib.rs:72 uses ANDROID_CONTEXT.expect("android context was not initialized").
The iroh 1.3.0 Endpoint Android documentation and iroh-dns 1.3.0 DnsResolver Android documentation explain that detecting an absent JNI context relies on panic unwinding and fails with panic = "abort". This matches the observed abort and profile. An unwind rebuild has not been performed, so this is a source-supported diagnosis, not yet a fix-validation claim.
Network controls
Termux Python socket.getaddrinfo("use1-1.relay.n0.iroh.link", 443) returned records. Python HTTPS requests to https://use1-1.relay.n0.iroh.link/ and https://github.com/n0-computer/sendme both returned HTTP 200 with normal certificate verification. This confirms ordinary DNS/HTTPS access, not QUIC/UDP reachability. Because --relay disabled also aborts immediately, no relay timeout is implicated by these tests.
Representative debug output (sanitized)
2026-11-09T20:42:37.250578Z DEBUG endpoint{id=<ENDPOINT ID>}: rustls::webpki::anchors: add_parsable_certificates processed 0 valid and 0 invalid certs
2026-11-09T20:42:37.250696Z DEBUG endpoint{id=<ENDPOINT ID>}: iroh::socket::transports::ip: binding addr=0.0.0.0:0
2026-11-09T20:42:37.250753Z TRACE endpoint{id=<ENDPOINT ID>}: mio::poll: registering event source with poller: token=Token(12970367443299711744), interests=READABLE | WRITABLE
2026-11-09T20:42:37.250734Z DEBUG endpoint{id=<ENDPOINT ID>}:portmapper.service: portmapper: portmap starting
2026-11-09T20:42:37.250788Z DEBUG endpoint{id=<ENDPOINT ID>}: iroh::socket::transports::ip: successfully bound addr=0.0.0.0:0 local_addr=0.0.0.0:37610
2026-11-09T20:42:37.250803Z DEBUG endpoint{id=<ENDPOINT ID>}: iroh::socket::transports::ip: binding addr=[::]:0
2026-11-09T20:42:37.250825Z TRACE endpoint{id=<ENDPOINT ID>}: mio::poll: registering event source with poller: token=Token(12970367443299714816), interests=READABLE | WRITABLE
2026-11-09T20:42:37.250844Z DEBUG endpoint{id=<ENDPOINT ID>}: iroh::socket::transports::ip: successfully bound addr=[::]:0 local_addr=[::]:49473
2026-11-09T20:42:37.250978Z TRACE endpoint{id=<ENDPOINT ID>}:portmapper.service: portmapper: tick: msg Some(UpdateLocalPort { local_port: Some(37610) })
RUST_BACKTRACE=full did not print a Rust stack to stderr for this stripped abort-profile binary. The Android crash buffer identifies the abort message and native PC offsets:
ABI: 'arm64'
signal 6 (SIGABRT), code -1 (SI_QUEUE), fault addr --------
Abort message: 'android context was not initialized'
backtrace:
#00 pc 00000000000599e4 /apex/com.android.runtime/lib64/bionic/libc.so (abort+164) (BuildId: c74277f481a383c87215b672f6465e24)
#01 pc 00000000007fc210 $HOME/.cargo/bin/sendme
#02 pc 00000000008171e8 $HOME/.cargo/bin/sendme
#03 pc 000000000081b698 $HOME/.cargo/bin/sendme
#04 pc 00000000006734b4 $HOME/.cargo/bin/sendme
#05 pc 000000000081a540 $HOME/.cargo/bin/sendme
#06 pc 000000000081a080 $HOME/.cargo/bin/sendme
#07 pc 000000000081b1c4 $HOME/.cargo/bin/sendme
#08 pc 000000000081b180 $HOME/.cargo/bin/sendme
#09 pc 000000000081b6b8 $HOME/.cargo/bin/sendme
#10 pc 00000000003420c4 $HOME/.cargo/bin/sendme
#11 pc 0000000000349708 $HOME/.cargo/bin/sendme
#12 pc 0000000000583d70 $HOME/.cargo/bin/sendme
#13 pc 0000000000819a60 $HOME/.cargo/bin/sendme
#14 pc 0000000000583c48 $HOME/.cargo/bin/sendme
#15 pc 000000000058b420 $HOME/.cargo/bin/sendme
#16 pc 000000000058ae64 $HOME/.cargo/bin/sendme
#17 pc 0000000000584a14 $HOME/.cargo/bin/sendme
#18 pc 00000000006936f0 $HOME/.cargo/bin/sendme
#19 pc 0000000000692178 $HOME/.cargo/bin/sendme
#20 pc 0000000000691c84 $HOME/.cargo/bin/sendme
#21 pc 000000000068e048 $HOME/.cargo/bin/sendme
#22 pc 00000000006a556c $HOME/.cargo/bin/sendme
#23 pc 0000000000825dcc $HOME/.cargo/bin/sendme
#24 pc 000000000082442c $HOME/.cargo/bin/sendme
#25 pc 0000000000830e1c $HOME/.cargo/bin/sendme
#26 pc 000000000082a0d0 $HOME/.cargo/bin/sendme
#27 pc 0000000000829eac $HOME/.cargo/bin/sendme
#28 pc 00000000007fd960 $HOME/.cargo/bin/sendme
#29 pc 0000000000804944 $HOME/.cargo/bin/sendme
#30 pc 00000000000be6a8 /apex/com.android.runtime/lib64/bionic/libc.so (__pthread_start(void*)+208) (BuildId: c74277f481a383c87215b672f6465e24)
#31 pc 000000000005b21c /apex/com.android.runtime/lib64/bionic/libc.so (__start_thread+64) (BuildId: c74277f481a383c87215b672f6465e24)
No unrelated applications' crash logs are included. All owned test processes were stopped; no user transfer was interrupted.
- Dominant language
- Rust
- Stars
- 1.3k
- Forks
- 83
- Avg merge
- 38m
- Merged PRs (30d)
- 2
Getting set up
This project ships no dev container, Dockerfile or contributing guide, so setting up is up to you: start from its README, and see our first-contribution guide for the general steps.
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
More from n0-computer/sendme
-
Stats with already finished downloadsPossibly taken @tahodev claimed this 28 days ago. Openready-for-pr
Difficulty 3/5 1-2 days Newbie friendliness 55/100
n0-computer/sendme#153 · 4 comments ·
Maintainers usually reply within 1 day
-
Difficulty 3/5 1-2 days Newbie friendliness 52/100
n0-computer/sendme#135 ·
Maintainers usually reply within 1 day
-
Difficulty 3/5 1-2 days Newbie friendliness 38/100
n0-computer/sendme#128 ·
Maintainers usually reply within 1 day
-
Difficulty 4/5 3-5 days Newbie friendliness 35/100
n0-computer/sendme#121 · 1 comment ·
Maintainers usually reply within 1 day
-
when receiving a large file, "exporting" takes significant additional time (few minutes)Possibly taken @mrsarthi claimed this 29 days ago. Openready-for-pr
Difficulty 3/5 1-2 days Newbie friendliness 48/100
n0-computer/sendme#120 · 2 comments · 2 reactions ·
Maintainers usually reply within 1 day
All issues in n0-computer/sendme
Similar issues
-
area:web kind:bug priority:p2 ready
Difficulty 2/5 1-3 hours Newbie friendliness 85/100
Maintainers usually reply within 1 day
-
bug
Difficulty 2/5 1-3 hours Newbie friendliness 80/100
Maintainers usually reply within 1 day
-
bug
Difficulty 1/5 Under an hour Newbie friendliness 84/100
sims1253/r-typeshed#140 · 1 comment ·
Maintainers usually reply within 1 day
-
Difficulty 2/5 1-3 hours Newbie friendliness 82/100
storytold/craft-libs#2867 ·
Maintainers usually reply within 1 day
-
help wanted needs investigation priority: low
Difficulty 2/5 1-3 hours Newbie friendliness 62/100