You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, I have found a bug as described in the title. The serve command hangs when a client attempts to open multiple connections and decides to send requests out-of-order, which actually seems be the behavior of real web browsers when dealing with multiple calls to the window.fetch(...) JS function. I can reproduce it with both the latest Firefox and Chromium.
This "httptest" repository is a simple reproducer - just clone it, serve it, and open the served index.html in a browser. It just runs the fetch() function on 10 files filled with zeroes. When everything loads, a paragraph should pop up saying Loaded all.. In the current main it stays on Loading... indefinitely.
Hi, I have found a bug as described in the title. The
serve
command hangs when a client attempts to open multiple connections and decides to send requests out-of-order, which actually seems be the behavior of real web browsers when dealing with multiple calls to thewindow.fetch(...)
JS function. I can reproduce it with both the latest Firefox and Chromium.This "httptest" repository is a simple reproducer - just clone it, serve it, and open the served
index.html
in a browser. It just runs thefetch()
function on 10 files filled with zeroes. When everything loads, a paragraph should pop up sayingLoaded all.
. In the currentmain
it stays onLoading...
indefinitely.I have a fix ready using
epoll
to make the TCPaccept
and the actual serving asynchronous relatively to each other, but that limits theserve
command to Linux, so it's probably not usable as-is? I can open a PR nevertheless, if desirable.The text was updated successfully, but these errors were encountered: