test: don't remove empty.txt on win32 by rvagg · Pull Request #2 · nodejs/node

rvagg · 2014-11-27T23:39:01Z

Take 3: on win32 we use empty.txt in the fixtures directory, otherwise we use a file constructed specifically for this test due to POSIX socket path length limitations, in which case we need to do appropriate cleanup. Cleaning up empty.txt in fixtures causes downstream problems for other tests using it (currently the next test is test-zlib).

on win32 we use empty.txt in the fixtures directory, otherwise we use a file constructed specifically for this test due to POSIX socket path length limitations, in which case we need to do appropriate cleanup

indutny · 2014-11-28T10:55:11Z

LGTM

bnoordhuis · 2014-11-28T16:51:42Z

Another LGTM.

bnoordhuis · 2014-11-28T16:52:55Z

@rvagg By the way, do you have commit rights? I'd like you to land your patches yourself.

rvagg · 2014-11-28T21:59:31Z

@bnoordhuis sure, is there anything special I need to know about the normal merge procedure?

bnoordhuis · 2014-11-29T00:44:11Z

@rvagg Just the normal commit log as described in CONTRIBUTING.md plus PR-URL and Reviewed-By tags. Basically, what you see when you run git log now.

on win32 we use empty.txt in the fixtures directory, otherwise we use a file constructed specifically for this test due to POSIX socket path length limitations, in which case we need to do appropriate cleanup Reviewed-By: Fedor Indutny <fedor@indutny.com> Reviewed-By: Ben Noordhuis <info@bnoordhuis.nl> PR-URL: #2

Release Candidate #2

siwenwang0803 · 2025-06-29T03:40:00Z

🤖 Secure-PR-Guard AI Code Review

Analysis completed at: 2025-06-28 20:39:59 UTC
PR: #2

🟡 Risk Assessment: MEDIUM

Total Issues Found: 4
Security Issues: 0

🟢 Low Priority Issues

Line 45 (length): Line 45 exceeds 120 characters.
Line 49 (length): Line 49 exceeds 120 characters.
Line 51 (length): Line 51 exceeds 120 characters.
Line 53 (length): Line 53 exceeds 120 characters.

🔗 Powered by Secure-PR-Guard | 🛡️ OWASP LLM Top 10 Compliant

This is an automated review. Please verify critical security findings manually.

siwenwang0803 · 2025-06-29T04:19:13Z

🤖 Secure-PR-Guard AI Code Review

Analysis completed at: 2025-06-28 21:19:12 UTC
PR: #2

🟡 Risk Assessment: MEDIUM

Total Issues Found: 2
Security Issues: 0

🟢 Low Priority Issues

Line 20 (length): Line exceeds 120 characters, consider breaking it into multiple lines for better readability.
Line 17 (style): Use of inconsistent indentation; consider using 2 spaces for proper code alignment instead of mixing styles.

🔗 Powered by Secure-PR-Guard | 🛡️ OWASP LLM Top 10 Compliant

This is an automated review. Please verify critical security findings manually.

…YErq3jenbmvHPj2ccYEnYH

This function call can fail with `Z_VERSION_ERROR` if the compiled library vs loaded library mismatched in version number or in stream structure size. In those cases, zlib doesn't initialize the `strm_.msg` field to null. Therefore, when a `CompressionError` object is created via `ErrorForMessage()`, it can read a stale or uninitialized `strm_.msg` pointer that will cause a crash. Example ASAN report: ``` AddressSanitizer: SEGV on unknown address #0 __strlen_avx2 string/../sysdeps/x86_64/multiarch/strlen-avx2.S:76 nodejs#1 strlen (/work/node/out/Debug/node+0x1a42ab7) nodejs#2 v8::(anonymous namespace)::StringLength(char const*) /work/node/out/../deps/v8/src/api/api.cc:7581:16 nodejs#3 v8::(anonymous namespace)::StringLength(unsigned char const*) /work/node/out/../deps/v8/src/api/api.cc:7587:10 nodejs#4 v8::String::NewFromOneByte(v8::Isolate*, unsigned char const*, v8::NewStringType, int) /work/node/out/../deps/v8/src/api/api.cc:7677:3 nodejs#5 node::OneByteString(v8::Isolate*, char const*, int, v8::NewStringType) /work/node/out/../src/util-inl.h:166:10 nodejs#6 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::EmitError(node::(anonymous namespace) ::CompressionError const&) /work/node/out/../src/node_zlib.cc:565:7 nodejs#7 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::CheckError() /work/node/out/../src/node_zlib.cc:519:5 nodejs#8 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::AfterThreadPoolWork(int) /work/node/out/../src/node_zlib.cc:543:10 nodejs#9 node::ThreadPoolWork::ScheduleWork() ::'lambda'(uv_work_s*, int) ::operator()(uv_work_s*, int) const /work/node/out/../src/threadpoolwork-inl.h:57:15 nodejs#10 node::ThreadPoolWork::ScheduleWork() ::'lambda'(uv_work_s*, int) ::__invoke(uv_work_s*, int) /work/node/out/../src/threadpoolwork-inl.h:48:7 nodejs#11 uv__work_done /work/libuv-1.51.0/src/threadpool.c:330:5 nodejs#12 uv__async_io.part.0 /work/libuv-1.51.0/src/unix/async.c:208:5 ``` Signed-off-by: ndossche <nora.dossche@ugent.be>

This function call can fail with `Z_VERSION_ERROR` if the compiled library vs loaded library mismatched in version number or in stream structure size. In those cases, zlib doesn't initialize the `strm_.msg` field to null. Therefore, when a `CompressionError` object is created via `ErrorForMessage()`, it can read a stale or uninitialized `strm_.msg` pointer that will cause a crash. Example ASAN report: ``` AddressSanitizer: SEGV on unknown address #0 __strlen_avx2 string/../sysdeps/x86_64/multiarch/strlen-avx2.S:76 #1 strlen (/work/node/out/Debug/node+0x1a42ab7) #2 v8::(anonymous namespace)::StringLength(char const*) /work/node/out/../deps/v8/src/api/api.cc:7581:16 #3 v8::(anonymous namespace)::StringLength(unsigned char const*) /work/node/out/../deps/v8/src/api/api.cc:7587:10 #4 v8::String::NewFromOneByte(v8::Isolate*, unsigned char const*, v8::NewStringType, int) /work/node/out/../deps/v8/src/api/api.cc:7677:3 #5 node::OneByteString(v8::Isolate*, char const*, int, v8::NewStringType) /work/node/out/../src/util-inl.h:166:10 #6 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::EmitError(node::(anonymous namespace) ::CompressionError const&) /work/node/out/../src/node_zlib.cc:565:7 #7 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::CheckError() /work/node/out/../src/node_zlib.cc:519:5 #8 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::AfterThreadPoolWork(int) /work/node/out/../src/node_zlib.cc:543:10 #9 node::ThreadPoolWork::ScheduleWork() ::'lambda'(uv_work_s*, int) ::operator()(uv_work_s*, int) const /work/node/out/../src/threadpoolwork-inl.h:57:15 #10 node::ThreadPoolWork::ScheduleWork() ::'lambda'(uv_work_s*, int) ::__invoke(uv_work_s*, int) /work/node/out/../src/threadpoolwork-inl.h:48:7 #11 uv__work_done /work/libuv-1.51.0/src/threadpool.c:330:5 #12 uv__async_io.part.0 /work/libuv-1.51.0/src/unix/async.c:208:5 ``` Signed-off-by: ndossche <nora.dossche@ugent.be> PR-URL: #63476 Reviewed-By: Anna Henningsen <anna@addaleax.net>

…ing fs.fsyncSync/fsync through module.exports so spies see it, ERR_INVALID_ARG_TYPE validation; node:test gets t.test subtests (inline, sequential, done-callback style), per-context t.mock with auto-restore, mock calls record error field; painpoints doc: nodejs#1/nodejs#2 marked resolved in milo compiler. fixes write-file-flush + append-file-flush, write/append cluster 13/13

This function call can fail with `Z_VERSION_ERROR` if the compiled library vs loaded library mismatched in version number or in stream structure size. In those cases, zlib doesn't initialize the `strm_.msg` field to null. Therefore, when a `CompressionError` object is created via `ErrorForMessage()`, it can read a stale or uninitialized `strm_.msg` pointer that will cause a crash. Example ASAN report: ``` AddressSanitizer: SEGV on unknown address #0 __strlen_avx2 string/../sysdeps/x86_64/multiarch/strlen-avx2.S:76 #1 strlen (/work/node/out/Debug/node+0x1a42ab7) #2 v8::(anonymous namespace)::StringLength(char const*) /work/node/out/../deps/v8/src/api/api.cc:7581:16 #3 v8::(anonymous namespace)::StringLength(unsigned char const*) /work/node/out/../deps/v8/src/api/api.cc:7587:10 #4 v8::String::NewFromOneByte(v8::Isolate*, unsigned char const*, v8::NewStringType, int) /work/node/out/../deps/v8/src/api/api.cc:7677:3 #5 node::OneByteString(v8::Isolate*, char const*, int, v8::NewStringType) /work/node/out/../src/util-inl.h:166:10 #6 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::EmitError(node::(anonymous namespace) ::CompressionError const&) /work/node/out/../src/node_zlib.cc:565:7 #7 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::CheckError() /work/node/out/../src/node_zlib.cc:519:5 #8 node::(anonymous namespace)::CompressionStream< node::(anonymous namespace)::ZlibContext> ::AfterThreadPoolWork(int) /work/node/out/../src/node_zlib.cc:543:10 #9 node::ThreadPoolWork::ScheduleWork() ::'lambda'(uv_work_s*, int) ::operator()(uv_work_s*, int) const /work/node/out/../src/threadpoolwork-inl.h:57:15 #10 node::ThreadPoolWork::ScheduleWork() ::'lambda'(uv_work_s*, int) ::__invoke(uv_work_s*, int) /work/node/out/../src/threadpoolwork-inl.h:48:7 #11 uv__work_done /work/libuv-1.51.0/src/threadpool.c:330:5 #12 uv__async_io.part.0 /work/libuv-1.51.0/src/unix/async.c:208:5 ``` Signed-off-by: ndossche <nora.dossche@ugent.be> PR-URL: #63476 Reviewed-By: Anna Henningsen <anna@addaleax.net>

…Context Code under vm.runInContext could read the whole host stack out of its own errors: `at Script.runInContext (node:vm:150:12)` and framework file paths sat right under the target's own frames, where a real browser realm shows nothing but its own. JS cannot close this. Error.prepareStackTrace has to stay undefined in web content or it is itself the tell, captureStackTrace only trims at construction, an Error constructor wrapper never sees the TypeErrors V8 throws itself, and lowering stackTraceLimit cuts the target's own frames too. V8 already does the right thing and Node turns it off. VisitStack drops frames whose native context holds a different security token than the context capturing the stack, and StackTrace::kDetailed does not carry kExposeFramesAcrossSecurityOrigins, so the filter is live for error.stack -- that is how one browser realm stays out of another's. node_contextify then hands every vm.Context the main context's token, and the filter never fires. So a boundary context is simply one holding its own token. The catch is that V8 access-checks every JSGlobalProxy and Isolate::MayAccess denies outright when the tokens differ and no callback is installed, which is why the token is shared in the first place. Contextify now installs an always-allow callback: contexts that share a token clear the check on the token comparison and never reach it, so nothing changes until a boundary is set, and vm was never a security boundary between host and sandbox. Cost of the extra check on the global proxy is under 1% (4e6 intercepted global accesses: 1714ms -> 1722ms), because the property interceptors already own that path. Because the filter runs while the stack is captured rather than while it is formatted, it covers errors V8 throws itself and target code that installs its own Error.prepareStackTrace to read CallSite objects directly. Measured on a context with the boundary on, the frames V8 hands out go from blob:x:inner | blob:x:outer | blob:x | node:vm | host.js | node:internal/... to blob:x:inner | blob:x:outer | blob:x The Error surface stays browser-shaped: stackTraceLimit 10, prepareStackTrace undefined, captureStackTrace a function. Ships as release nodejs#2, "She was more like a beauty queen from a movie scene" (Billie Jean). The clock-surface test now pins the shape of the release record instead of the words, so a new lyric no longer edits a test. Requested by the rex-mirror-realm worker semantics work, where the two probes still red are "no node: internal frames in the stack" and "no host disk paths in the stack". Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_013ng1GuMcCzJEFDAAEmujWC

test: don't remove empty.txt on win32

9006e94

on win32 we use empty.txt in the fixtures directory, otherwise we use a file constructed specifically for this test due to POSIX socket path length limitations, in which case we need to do appropriate cleanup

rvagg merged this pull request into v0.12 Nov 29, 2014

rvagg deleted the test-net-pipe-connect-errors-socket-path-win-fix-2 branch November 29, 2014 01:00

jmar777 mentioned this pull request Feb 21, 2015

Feature Request: Every async function returns Promise #11

Closed

JediMindtrick mentioned this pull request Apr 4, 2015

io.js community thoughts on reconciliation #1336

Closed

yjhjstz mentioned this pull request May 14, 2015

DebuggerAgent: unhandled socket error event #781

Closed

dashed mentioned this pull request May 31, 2015

a more deliberate release process #1854

Closed

Fishrock123 mentioned this pull request Jun 21, 2015

test: regression for fs module monkey-patching #2026

Closed

snorp mentioned this pull request Aug 7, 2015

Timeouts sometimes not fired on Linux/ARM #2326

Closed

silverwind mentioned this pull request Sep 23, 2015

doc: jenkins-iojs.nodesource.com -> ci.nodejs.org, iojs.org/download -> nodejs.org/download #2886

Closed

jasnell mentioned this pull request Oct 12, 2015

Release procedure notes #3254

Closed

silverwind mentioned this pull request Oct 14, 2015

Consistent error messages in all modules #3374

Closed

kvic-z mentioned this pull request Oct 27, 2015

compile error ,target=arm7 #3224

Closed

jasnell added a commit that referenced this pull request Nov 2, 2015

2015-10-29, Version 4.2.2 RC2 (LTS)

dd1f411

Release Candidate #2

jasnell mentioned this pull request Nov 16, 2015

doc: add link to [customizing util.inspect colors]. #3749

Closed

chvostek mentioned this pull request Nov 20, 2015

Is it possible to provide binary for FreeBSD #3716

Closed

gireeshpunathil mentioned this pull request Nov 24, 2015

Segfault in node::StreamBase::GetFD #2847

Closed

alanpurple mentioned this pull request Dec 8, 2015

v8 4.6.85.31 typed array bug #4191

Closed

jianpingw mentioned this pull request Dec 13, 2015

node js 5.2.0 segmentation fault after console.log #4261

Closed

juanarbol mentioned this pull request Mar 3, 2023

V18.15.0 proposal #46920

Merged

mshopf mentioned this pull request Jun 15, 2023

Closing filehandle after being used in a stream exits node without error message #48466

Closed

khagankhan mentioned this pull request Oct 13, 2023

Use of Uninitialized Memory Detected in bytecode_builtins_list_generator by MemorySanitizer #50171

Closed

orsett0 mentioned this pull request Dec 28, 2023

Segmentation fault when allocating memory in WASM #51303

Closed

bocchi810 mentioned this pull request Jan 4, 2024

Can't build on gcc-g++ 14.0.0 #51369

Closed

MylesBorins mentioned this pull request Jan 10, 2024

enable corepack by default #50963

Closed

kshaa mentioned this pull request Mar 1, 2024

Stream memory leak in 20.11.1 #51936

Closed

eugeneo mentioned this pull request Apr 15, 2024

src: fix Worker termination in inspector.waitForDebugger #52527

Merged

benjamingr mentioned this pull request Jun 28, 2024

Nominating @MattiasBuelens as collaborator #52831

Closed

avivkeller mentioned this pull request Jul 6, 2024

doc: fix that 22.x docs scroll to wrong locations #53662

Merged

Qard mentioned this pull request Feb 3, 2025

src: fix platform shutdown deadlock #56827

Closed

ronking-zhu mentioned this pull request Oct 13, 2025

Nodejs official released build has conflict with msodbcsql driver due to incompatible openssl version #60240

Closed

Mohit-Raj-Saxena mentioned this pull request Oct 16, 2025

Multiple Documentation Issues in fs and fs/promises Modules (Formatting, Consistency, and Clarity) #60280

Closed

sadan4 mentioned this pull request Nov 5, 2025

NodeJS segfaults when trying to debug this code #60580

Closed

jackzhp mentioned this pull request Nov 5, 2025

Large file download (above 50MB, without "range processing") results in HTTP2_PROTOCOL_ERROR in Chrome #59465

Closed

nabsul added a commit to nabsul/node that referenced this pull request Nov 18, 2025

Try nodejs#2

42d0f18

Copilot AI referenced this pull request in Multi-V-VM/node-wasix32 Mar 20, 2026

Merge pull request #2 from tonidy/claude/fix-v8-timedeltanamespace-01…

9f9e4e2

…YErq3jenbmvHPj2ccYEnYH

alexandercarlis2-dotcom mentioned this pull request Jun 13, 2026

Node.js API #63896

Closed

Navigation Menu

Uh oh!

Uh oh!

test: don't remove empty.txt on win32 - #2

test: don't remove empty.txt on win32#2
rvagg merged 1 commit into
v0.12from
test-net-pipe-connect-errors-socket-path-win-fix-2

rvagg commented Nov 27, 2014

Uh oh!

indutny commented Nov 28, 2014

Uh oh!

bnoordhuis commented Nov 28, 2014

Uh oh!

bnoordhuis commented Nov 28, 2014

Uh oh!

rvagg commented Nov 28, 2014

Uh oh!

bnoordhuis commented Nov 29, 2014

Uh oh!

siwenwang0803 commented Jun 29, 2025

Uh oh!

siwenwang0803 commented Jun 29, 2025

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Uh oh!

Uh oh!

Conversation

rvagg commented Nov 27, 2014

Uh oh!

indutny commented Nov 28, 2014

Uh oh!

bnoordhuis commented Nov 28, 2014

Uh oh!

bnoordhuis commented Nov 28, 2014

Uh oh!

rvagg commented Nov 28, 2014

Uh oh!

bnoordhuis commented Nov 29, 2014

Uh oh!

siwenwang0803 commented Jun 29, 2025

🤖 Secure-PR-Guard AI Code Review

🟡 Risk Assessment: MEDIUM

🟢 Low Priority Issues

Uh oh!

siwenwang0803 commented Jun 29, 2025

🤖 Secure-PR-Guard AI Code Review

🟡 Risk Assessment: MEDIUM

🟢 Low Priority Issues

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants