profile
viewpoint
Johan Bergström jbergstroem Netwrk Santiago, Chile https://bergstroem.nu Content optimiser by day, Node.js infra by night. DJ early mornings and Tetris player in between.

chriso/bitset 15

A compressed bitset with supporting data structures and algorithms

jbergstroem/atom-icon 5

An atom icon replacement. And a way to install it.

jbergstroem/build 1

my fork of the build repo. be warned!

jbergstroem/cloudflared 1

Argo Tunnel client

jbergstroem/docker-node 1

Official Docker Image for Node.js

jbergstroem/gentoo-overlay 1

My gentoo overlay

jbergstroem/aports 0

Mirror of aports repository

jbergstroem/bankai 0

:station: - The easiest way to compile JavaScript, HTML and CSS

jbergstroem/bcrypt.js 0

Optimized bcrypt in plain JavaScript with zero dependencies.

jbergstroem/bhyveucl 0

Script for starting bhyve instances based on a libUCL config file

created repositoryrvagg/js-bitcoin

JavaScript Bitcoin data multiformats codecs and utilities for IPLD

created time in 2 minutes

push eventlandley/toybox

Eric Molitor

commit sha db0737648cccfcc8dc29479460c10f296fbb052a

Merge ipv6 setroute into a single function

view details

Eric Molitor

commit sha 52fa7bf4fca2afe8853225d674b8ac7e036b315f

Remove unused include and constants

view details

push time in 3 hours

issue closedlandley/toybox

0.8.2+ Fails to build on mips

I started to see this error on mips platform with (gcc10/glibc-2.31),

lib/portability.c:404:21: error: 'SIGSTKFLT' undeclared here (not in a function); did you mean 'SIGSTKSZ'?
  404 | #define SIGNIFY(x) {SIG##x, #x}
      |                     ^~~
lib/portability.c:416:3: note: in expansion of macro 'SIGNIFY'
  416 |   SIGNIFY(STKFLT), SIGNIFY(POLL), SIGNIFY(PWR),
      |   ^~~~~~~

detailed log here

closed time in 3 hours

kraj

issue commentlandley/toybox

0.8.2+ Fails to build on mips

Committed in 824663599378

kraj

comment created time in 3 hours

push eventlandley/toybox

Rob Landley

commit sha 824663599378774ebbb2101df60197f9eb847910

Instead of ifdef linux or macos, just ifdef the signals.

view details

push time in 3 hours

pull request commentlandley/toybox

lib.[ch]: switch loggit to macro, to avoid non-standard vsyslog

Related question, is there a reason to keep loggit in lib, given that it's only used in su?

E5ten

comment created time in 4 hours

PR opened landley/toybox

dhcpd: fix warnings

Add presumably intended conversion specifier for gconfig.valid_lifetime as it is an argument to error_msg despite the format string having no conversion specifiers. Remove always true false inverse check for address of array gstate.rcvd.rcvd_pkt6.transaction_id.

+2 -3

0 comment

1 changed file

pr created time in 6 hours

pull request commentlandley/toybox

Replace [gs]ettimeofday with clock_[gs]ettime

In addition to musl ignoring the settimeofday timezone argument, mentioned in the hwclock commit, glibc 2.31 and up apparently errors when settimeofday is called with both arguments non-NULL.

E5ten

comment created time in 8 hours

issue closedh2o/h2o

test failures when running make check

Trying to get to consistently successful make check runs on Alpine Linux (latest stable 3.11). Reason being so that the check target can be run as part of package build.

It maybe worthwhile to focus that goal on the latest, assuming there is a plan for a release, but I am stuck with even building master on Alpine.

For that reason executing it on v2.2.6 and v2.3.0-beta2, but tests are failing. To get the check target working I had to apply attached patches, which are issues that have been addressed, but not back ported to v2.2 respectively 2.3.0-betaX. But then I have test failures - see attachments. Any help to work through those would be great. Also, if there is a slimmed down check target that maybe is sufficient enough to run that proves the core functionality is there, that is not obvious, that maybe an alternative approach rather than getting ALL tests to run successfully.

cmake invocation and output (v2.3.0-beta2):

$ cmake -B build . -DCMAKE_BUILD_TYPE=None -DBUILD_SHARED_LIBS=ON -DCMAKE_INSTALL_LIBDIR=/usr/lib -DCMAKE_INSTALL_PREFIX=/usr -DWITH_MRUBY=ON
-- The C compiler identification is GNU 9.2.0
-- The CXX compiler identification is GNU 9.2.0
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Found PkgConfig: /usr/bin/pkg-config (found version "1.6.3") 
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD
-- Performing Test CMAKE_HAVE_LIBC_PTHREAD - Success
-- Found Threads: TRUE  
-- Found OpenSSL: /usr/lib/libcrypto.so (found version "1.1.1g")  
-- Found ZLIB: /lib/libz.so (found version "1.2.11") 
-- Performing Test ARCH_SUPPORTS_64BIT_ATOMICS
-- Performing Test ARCH_SUPPORTS_64BIT_ATOMICS - Success
-- Performing Test HAS_PTHREAD_SETAFFINITY_NP
-- Performing Test HAS_PTHREAD_SETAFFINITY_NP - Success
-- Performing Test LIBC_HAS_BACKTRACE
-- Performing Test LIBC_HAS_BACKTRACE - Success
-- Checking for module 'libuv>=1.0.0'
--   Found libuv, version 1.34.0
-- Checking for module 'libwslay'
--   Found libwslay, version 1.1.0
-- Configuring done
-- Generating done
-- Build files have been written to: /home/axel/h2o/build

h2o-v226-makecheck.log h2o-v230beta2-makecheck.log h2o-v230beta2.patch.txt h2o-v226.patch.txt

closed time in 8 hours

axelu

issue commenth2o/h2o

test failures when running make check

Worked through all issues I had on 2.2.6 and 2.3.0-beta2 getting make check to work on Alpine, some resulting in Issues and PRs.

PRs

  • #2330
  • #2331
  • #2334

Issues

  • #2329
axelu

comment created time in 8 hours

PR opened landley/toybox

Remove XSI putenv usage, replace with POSIX setenv
+6 -6

0 comment

2 changed files

pr created time in 8 hours

PR opened landley/toybox

Replace [gs]ettimeofday with clock_[gs]ettime

gettimeofday is XSI, not POSIX, and is marked obsolete, while settimeofday is non-standard. Switch use of these functions to their modern replacements.

+134 -145

0 comment

6 changed files

pr created time in 11 hours

issue commentlandley/toybox

0.8.2+ Fails to build on mips

at the moment we have SIGEMT in an __APPLE__ ifdef. but it turns out there's a SIGEMT on mips linux too. so it should move out of the __APPLE__ and just be #ifdef SIGEMT instead.

kraj

comment created time in 14 hours

startedzwegner/x86-info-term

started time in 18 hours

push eventlandley/toybox

Rob Landley

commit sha bc6ce6628c2517aa419b5551b8d378da794b84f2

More sh tests.

view details

Rob Landley

commit sha 6b6436c848035abfff92a86b00305847bc6c0b9d

Fix syntax checking for "if true; then echo hello | fi", fix some tests, slightly more elaborate debug output.

view details

push time in 19 hours

pull request commentnodejs/security-wg

Onboard Eva Sarafianou to the WG

@sam-github @mhdawson can you add @esarafianou to the @nodejs/security-wg team?

esarafianou

comment created time in 21 hours

issue closednodejs/security-wg

Nominate Eva Sarafianou to the triage team

Hi @nodejs/security-wg

I would like to nominate Eva Sarafianou (@esarafianou) to the triage team.

Eva is a product security engineer at Auth0. She brings tons of experience with vulnerability management and JavaScript security and I am confident she will be a great asset to the team.

I can onboard Eva to our processes.

closed time in a day

MarcinHoppe

PR merged nodejs/security-wg

Reviewers
Onboard Eva Sarafianou to the WG member

Adding myself to the members; fixes #657

+3 -1

3 comments

2 changed files

esarafianou

pr closed time in a day

push eventnodejs/security-wg

Eva Sarafianou

commit sha 77654b8db17ea7deeabd33073af360b0fc906a5f

Onboard esarafianou to the WG (#659)

view details

push time in a day

push eventlandley/toybox

Eric Molitor

commit sha d565cf631cdd5bdeaa7be00be83516bb649446a1

Use perror_exit to replace exit calls.

view details

Eric Molitor

commit sha 729e3081f6e630f17a5d0027bdfda91131d65125

Remove send_nlrtmsg embedding call into display_route as it is only used here

view details

Eric Molitor

commit sha d3f83232d49e26f53121fc1f797118080a0fdf73

First pass of adding RTNetlink setroute support

view details

push time in a day

PublicEvent

push eventlandley/toybox

Rob Landley

commit sha b79f2e8c6e1fe5c6c518bf801cbc5a7eef06e381

Tweak roadmap so status.html is a bit more current.

view details

push time in a day

issue commentlandley/toybox

echo and \0

Analyzed stuff a bit more and there's a bug for sscanf in unescape2 and because of this, the off jumps past the end of the string and continues printing junk until it hits another real string terminator.

The check needs to be changed to if (sscanf(...) > 0) as a an ending \0 will attempt to trigger on the octal string, but it returns EOF because it didn't match before reaching end of input. Because the return value is not checked for errors, the EOF (-1) triggers the if and we jump past the end of the string and hell ensues.

dlegaultbbry

comment created time in a day

issue closedh2o/h2o

startup delay

after h2o is started, it takes minutes before it is ready to serve requests. The error log reports the start of the worker instantly (1st line in below log excerpt), and then minutes later it adds the lines as shown below, and only then is it serving requests. It is listening instantly after restart, meaning the ports are open, but it is not answering. It applies to both ports, http and https. There are no other log entries related to this anywhere else on the system. I assume it has to do with the OCSP stapling, but not sure. Any help to get it to answer faster after a restart is appreciated, including help how to debug what it is actually doing respectively waiting for during that time. Issue observed in 2.2.6 and 2.3.0_beta1.

starting new worker 1109 [INFO] raised RLIMIT_NOFILE to 1073741816 h2o server (pid:1109) is ready to serve requests fetch-ocsp-response (using OpenSSL 1.1.1d FIPS 10 Sep 2019) fetch-ocsp-response (using OpenSSL 1.1.1d FIPS 10 Sep 2019) sending OCSP request to http://ocsp.int-x3.letsencrypt.org sending OCSP request to http://ocsp.int-x3.letsencrypt.org /etc/letsencrypt/live/www.domain.com/fullchain.pem: good This Update: Apr 25 13:00:00 2020 GMT Next Update: May 2 13:00:00 2020 GMT verifying the response signature /etc/letsencrypt/live/www.domain.com/fullchain.pem: good This Update: Apr 25 13:00:00 2020 GMT Next Update: May 2 13:00:00 2020 GMT verifying the response signature verify OK (used: -VAfile /tmp/Hhf6gP1kzN/issuer.crt) [OCSP Stapling] successfully updated the response for certificate file:/etc/letsencrypt/live/www.domain.com/fullchain.pem verify OK (used: -VAfile /tmp/eV54DySy1h/issuer.crt) [OCSP Stapling] successfully updated the response for certificate file:/etc/letsencrypt/live/www.domain.com/fullchain.pem

closed time in 2 days

axelu

issue commenth2o/h2o

startup delay

@paulpref Thank you for sharing. Issue for me was on Fedora 31, h2o started via systemd with LimitNOFILE=infinity in the corresponding service file. That is in fact the issue. Thank you again.

axelu

comment created time in 2 days

push eventh2o/h2o

Kazuho Oku

commit sha 6d91ec01b791196c4ac5df8cd4238b55f5ae7231

Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "link picotest so that picotls can be built""""""""""""""""""""""""""""""""""""""" This reverts commit b85efc6d128ace04a5c0bc3f8c86a0a020ef57ee.

view details

Kazuho Oku

commit sha 89867e0177027731328e23feac5f714d2de1e036

extract https://github.com/h2o/picotls @ 89f71e9 () at deps/picotls

view details

Kazuho Oku

commit sha c3024088014e847f1ae32c909e4a3fdd31ee3b9a

Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "Revert "link picotest so that picotls can be built"""""""""""""""""""""""""""""""""""""""" This reverts commit 6d91ec01b791196c4ac5df8cd4238b55f5ae7231.

view details

Kazuho Oku

commit sha ea8a9099f8e9d38b1ad938ea85b03912c3126d5e

update comment

view details

Kazuho Oku

commit sha f423c0207d584d4f6c16a369fb3c336dbb9ed6e2

Merge pull request #2335 from h2o/kazuho/update-picotls-5 update picotls

view details

push time in 2 days

PR merged h2o/h2o

update picotls
+238 -100

0 comment

8 changed files

kazuho

pr closed time in 2 days

PR opened h2o/h2o

update picotls
+238 -100

0 comment

8 changed files

pr created time in 2 days

more