Lighttpd

2020-11-14

01:54 Feature #1982: RFE: mod_redirect exact-match map: simple (non-RE) redirection

[Title (subject) renamed to provide some scope.]
Should anyone find this and wish to try extending mod_redirect or c... gstrauss

2020-11-13

22:00 Revision c42b2805: [core] fix bug when HTTP/2 frames span chunks

fix inverted logic when HTTP/2 frames span chunkqueue chunks gstrauss

08:51 Bug #3033: Memory Growth with PUT and full buffered streams

(issue automatically closed with patch push)
If this issue is still unresolved, please reopen. If a new or differ... gstrauss

08:50 Bug #3033 (Fixed): Memory Growth with PUT and full buffered streams

Applied in changeset commit:f2b33e752009065f85611a71c18ac5df94247ce7. gstrauss

06:40 Bug #3033: Memory Growth with PUT and full buffered streams

@flynn: I have seen clean behavior from mod_proxy as I tried to reverse-proxy a 256 MB PUT.
If I understand correc... gstrauss

2020-11-12

22:41 Bug #3033 (Patch Pending): Memory Growth with PUT and full buffered streams

I pushed a small patch to the tip of my dev branch @personal/gstrauss/master@. This should address the potential lar... gstrauss

10:48 Bug #3033: Memory Growth with PUT and full buffered streams

I applied your patch and tested three iterations with upload of 350MB, still only 5MB private dirty memory (before th... flynn

10:17 Bug #3033: Memory Growth with PUT and full buffered streams

At the very least, I need to reschedule the request once the queue to the backend clears, in the case where we have a... gstrauss

10:07 Bug #3033: Memory Growth with PUT and full buffered streams

I need to verify this further, as this might be an incomplete patch, but the issue appears to be in mod_fastcgi aggre... gstrauss

09:56 Bug #3033: Memory Growth with PUT and full buffered streams

I can reproduce the problem with your up.fcgi backend and a 900MB file upload.
I use always the current Debian pac... flynn

09:25 Bug #3033: Memory Growth with PUT and full buffered streams

With larger files, I do see more memory usage, so I have something to look into and instrument. More tomorrow. gstrauss

09:20 Bug #3033: Memory Growth with PUT and full buffered streams

Tested my config above adding @mod_openssl@, and, separately, adding @mod_mbedtls@. Testing over TLS used more memor... gstrauss

09:13 Bug #3033: Memory Growth with PUT and full buffered streams

Same behaviout without TLS connection: the memory steps for each iteration are different, the memory amount after fou... flynn

09:07 Bug #3033: Memory Growth with PUT and full buffered streams

> I know at least two clients in use, which do PUT requests without Content-Length, so it is important for me.
No ... gstrauss

08:27 Bug #3033: Memory Growth with PUT and full buffered streams

FYI: we already have had a discussion about PUT and chunked encoding, see ticket #2156. You fixed it, but only if set... flynn

08:19 Bug #3033: Memory Growth with PUT and full buffered streams

I did not change/configure ssl.read-ahead.
The connections are all over TLS, I try to make tests without ... flynn

08:12 Bug #3033: Memory Growth with PUT and full buffered streams

Something is clearly wrong in lighttpd for it to use that much memory from sequential PUT requests which ought to be ... gstrauss

07:52 Bug #3033: Memory Growth with PUT and full buffered streams

I use big files to trigger the memory leak, currently I use a debian netinst iso image with 350MB.
The first uploa... flynn

06:54 Bug #3033: Memory Growth with PUT and full buffered streams

Trying to reproduce. Unable to reproduce using basic CGI
@lighttpd.conf@... gstrauss

22:19 Revision f2b33e75: [multiple] add back-pressure gw data pump (fixes #3033)

When server.stream-request-body = 0 (the default), the entire request
body is collected before engaging the backend. ... gstrauss

06:44 Revision 45aa1aa8: [mod_cgi] ensure tmp file open() before splice()

(bug on master branch)
With lighttpd defaults, including fully buffering request body, and
if request body > 1 MB, t... gstrauss

06:43 Revision fc19558f: [mod_cgi] fix crash if initial write to CGI fails

(bug on master branch) gstrauss

01:57 Revision 78ec2b5b: [core] filter out duplicate modules

filter out modules duplicated in server.modules list gstrauss

00:06 Revision c16c6a8f: [mod_alias] validate given order, not sorted order

(bug on master branch)
x-ref:
"Debian Bullseye/sid arm64 - lighttp broken after update"
https://discussions.flig... gstrauss

2020-11-11

22:02 Bug #3033: Memory Growth with PUT and full buffered streams

- currently I do not use valgrind
- I use the processes plugin of collectd and a self written plugin which sums up th... flynn

21:00 Bug #3033: Memory Growth with PUT and full buffered streams

Does a program such as valgrind report the memory as leaked? Or are you merely seeing the lighttpd memory footprint ... gstrauss

20:34 Bug #3033 (Fixed): Memory Growth with PUT and full buffered streams

My memory monitoring for lighttpd shows, that some lighttpd instances leak memory.
I could track it down to PUT-re... flynn

19:57 Feature #3032 (Wontfix): Add a privacy option to allow IP Anonymization when creating logfiles.

FYI: You got the response you got because you *failed* to include basic information about requirements (GDPR) with yo... gstrauss

10:13 Feature #3032: Add a privacy option to allow IP Anonymization when creating logfiles.

I am aware of Option 1, but I do not want to omit the IP in full.
I did not yet found a piped logger that rewrites... dinoex

08:16 Revision 82273a94: [doc] update optional pkg dependencies in INSTALL

gstrauss

2020-11-10

22:59 Feature #3032 (Need Feedback): Add a privacy option to allow IP Anonymization when creating logfiles.

Why?
See [[Docs_ModAccessLog|mod_accesslog]] docs. You can already
1. define a log format which omits the IP or use... gstrauss

21:24 Feature #3032 (Wontfix): Add a privacy option to allow IP Anonymization when creating logfiles.

With IPv4:
IP Anonymization clears the last 8 Bit of an IP Address
With IPv6:
IP Anonymization clears the last 6... dinoex

12:44 Revision 5ec13918: [core] define SHA*_DIGEST_LENGTH macros if missing

gstrauss

11:10 Revision 47aa6d4a: [core] http_response_buffer_append_authority()

make public func for benefit of external, third-party mod_authn_tkt gstrauss

07:33 Bug #3030: Fastcgi fails if server.tag is empty

I didn't fix the one place you needed it! Sorry.
I applied a reverse patch from yours. gstrauss

06:32 Bug #3030: Fastcgi fails if server.tag is empty

Thank for your quick response and fix.
Tested and works for fastcgi/php, but uwsgi is missing, see my attached patch. flynn

00:05 Bug #3030 (Fixed): Fastcgi fails if server.tag is empty

Applied in changeset commit:c0e2667b714bf8be8d5cb2da44a27a25a68212be. gstrauss

07:32 Revision 291fd1e7: [core] accept "HTTP/2.0", "HTTP/3.0" from backends (#3031)

accept "HTTP/2.0" and "HTTP/3.0" NPH from naive non-proxy backends
(thx flynn)
x-ref:
"uwsgi fails with HTTP/2"
... gstrauss

06:32 Bug #3031: uwsgi fails with HTTP/2

Tested and works, but only with none empty server.tag, see my remark on #3030. flynn

00:05 Bug #3031 (Fixed): uwsgi fails with HTTP/2

Applied in changeset commit:169d8d3608fb90e4f00ec77d149788e59a6fc8d5. gstrauss

00:01 Bug #3031 (Patch Pending): uwsgi fails with HTTP/2

gstrauss

01:26 Revision 5e711068: [build] check for xxhash in more ways

pkg-config libxxhash.pc might not be provided with xxhash < 0.7.3
x-ref:
"Update build-dep for xxhash [...]"
htt... gstrauss

00:00 Revision 169d8d36: [core] accept "HTTP/2.0", "HTTP/3.0" from backends (fixes #3031)

accept "HTTP/2.0" and "HTTP/3.0" NPH from naive non-proxy backends
(thx flynn)
x-ref:
"uwsgi fails with HTTP/2"
... gstrauss

2020-11-09

23:20 Revision c0e2667b: [multiple] handle NULL val as empty in *_env_add (fixes #3030)

(bug on master branch; never released)
(thx flynn)
x-ref:
"Fastcgi fails if server.tag is empty"
https://redmin... gstrauss

22:30 Bug #3031: uwsgi fails with HTTP/2

Thanks for reporting this. This is a bug in uwsgi, but I'll think about a workaround in lighttpd.
Protocols that fo... gstrauss

14:12 Bug #3031 (Fixed): uwsgi fails with HTTP/2

Updating to the upcoming version 1.4.56 and enabling HTTP/2 my searx instance fails with a 502.
Using strace I fou... flynn

22:24 Bug #3030 (Patch Pending): Fastcgi fails if server.tag is empty

Thanks for tracking that down.
I've modified the code to accept a 0-length value with NULL ptr when creating the C... gstrauss

13:05 Bug #3030 (Fixed): Fastcgi fails if server.tag is empty

Updating to the upcoming version 1.4.56 all my php-Script fail with a return code 400.
Reducing my setup I found,... flynn

10:07 Revision 21337394: [tests] allow LIGHTTPD_EXE_PATH override

allow LIGHTTPD_EXE_PATH override to be able to run source tree tests/*.t
against installed executable, e.g. LIGHTTPD_... gstrauss