MantisBT - Issues

0002897: Unable to set minimum font size

Tue, 10 Mar 2026 10:32:40 +0000

Documentation states that minimum font size can be set and displayed font will never be smaller than it. The option is missing from 'Content' configuration tag. I found a command line switch --font_min_size=. It is accepted, but quietly ignored.

0002896: Error link web interface.

Fri, 20 Feb 2026 22:31:42 +0000

Error link on web interface on Source control Page. Whern clicked go to 404

0002895: SVG issues

Mon, 15 Dec 2025 08:05:03 +0000

Not sure if this is a bug or just incomplete development. If the latter I apologise and will await the next update. However would someone clarify the progress / Image support / SVG page please?

SVG-only file (of file type SVG) dragged to Netsurf's icon bar icon will render happily, but trying to do so as part of an HTML page (inline or file-refereneced) results in only the HTML being displayed.

0002894: BBC News front page: complete mess

Thu, 11 Dec 2025 15:25:00 +0000

Visit https://www.bbc.co.uk/news and you'll see a mess. The top part of the page has items superimposed, and repeats 5 times as you scroll down.

0002893: Embedded SVG support

Sun, 30 Nov 2025 23:06:00 +0000

Despite already having proper SVG support for images linked normally via the img src=... tag, there is still a lack of support for embedded SVG.

SVG in HTML is not something uncommon, it's used on many sites. It would be great to have support for inline SVG.

0002892: multipe home page url's no accepted

Sat, 01 Mar 2025 01:55:53 +0000

https://www.abc.net.au/newengland%20%20|%20%20https://www.abc.net.au/brokenhill%20|%20https://www.nine.com.au/sport%20|%20https://new.reddit.com/

....The page you are looking for cannot be found. It might have been removed, had its name changed, or be temporarily unavailable.

I would usually enter a url....and then separate it from the next url with a |

0002891: build error - cannot read xpm

Wed, 26 Feb 2025 06:34:24 +0000

downloaded v3.11 tar sources
set gtk to version 3
run make command

COMPILE: build/Linux-gtk3/duktape/xml_serializer.c
MKDIR: build/Linux-gtk3/tools
BUILD CC: build/Linux-gtk3/tools/split-messages
MSGSPLIT: Language: de Filter: gtk
MSGSPLIT: Language: en Filter: gtk
MSGSPLIT: Language: fr Filter: gtk
MSGSPLIT: Language: it Filter: gtk
MSGSPLIT: Language: nl Filter: gtk
MSGSPLIT: Language: zh_CN Filter: gtk
GRESORCE: frontends/gtk/res/messages.gresource.xml
COMPILE: build/Linux-gtk3/messages_gresource.c
GRESORCE: frontends/gtk/res/netsurf.gresource.xml
failed to load "frontends/gtk/res/netsurf.xpm": Couldn?t recognize the image file format for file ?frontends/gtk/res/netsurf.xpm?
frontends/gtk/res/netsurf.gresource.xml: Child process exited with code 1.
make[1]: *** [frontends/gtk/Makefile:114: build/Linux-gtk3/netsurf_gresource.c] Error 1
make[1]: Leaving directory '/home/user/Downloads/netsurf-all-3.11/netsurf'
make: *** [Makefile:125: /home/user/Downloads/netsurf-all-3.11/inst-gtk3/build-stamp] Error 2
user@Personal2:~/Downloads/netsurf-all-3.11$

0002890: libdom crashes when built with libxml2-2.13.x

Sat, 01 Mar 2025 18:10:46 +0000

Apparently libdom is relying on some internal details of libxml2 during parsing (bindings/xml/libxml_xmlparser.c). The first indication of this is that the libdom test suite fails after upgrading to libxml2-2.13.5. Upon closer inspection, libdom is segfaulting because there's a NULL where it isn't expecting one (missing parent).

In short, the problem is with the assumption that this will retrieve the parent element:

xmlNodePtr parent = parser->xml_ctx->node;

In libxml2-2.13.x, XML entities are parsed into a dummy element, and we are retrieving that dummy element with the code above. In earlier versions of libxml2, the same entities were parsed into a temporary context where xml_ctx->node by coincidence still points to the previous node, so it happened to work.

Refs:

* https://bugs.gentoo.org/946980
* https://gitlab.gnome.org/GNOME/libxml2/-/issues/837

0002887: make check fails: fail_unless has too many arguments for format

Thu, 09 Jan 2025 15:03:01 +0000

make test for libwapcaplet fails with the following errors, when compiles with -Werror=format-extra-args:

make -j2 test V=1 PREFIX=/usr COMPONENT_TYPE=lib-shared
make[2]: Entering directory '/builds/aelmahmoudy/libwapcaplet/debian/output/source_dir'
COMPILE: test/testmain.c
COMPILE: test/basictests.c
In file included from test/basictests.c:9:
test/basictests.c: In function 'test_lwc_intern_string_ok_fn':
test/basictests.c:119:21: error: too many arguments for format [-Werror=format-extra-args]
119 | "Unable to intern a simple string");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:121:21: error: too many arguments for format [-Werror=format-extra-args]
121 | "Returned OK but str was still NULL");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_intern_string_twice_ok_fn':
test/basictests.c:129:21: error: too many arguments for format [-Werror=format-extra-args]
129 | "Unable to intern a simple string");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:131:21: error: too many arguments for format [-Werror=format-extra-args]
131 | "Returned OK but str was still NULL");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:133:21: error: too many arguments for format [-Werror=format-extra-args]
133 | "Unable to intern a simple string");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:135:21: error: too many arguments for format [-Werror=format-extra-args]
135 | "Returned OK but str was still NULL");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_intern_string_twice_same_ok_fn':
test/basictests.c:143:21: error: too many arguments for format [-Werror=format-extra-args]
143 | "Unable to intern a simple string");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:145:21: error: too many arguments for format [-Werror=format-extra-args]
145 | "Returned OK but str was still NULL");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:147:21: error: too many arguments for format [-Werror=format-extra-args]
147 | "Unable to intern a simple string");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:149:21: error: too many arguments for format [-Werror=format-extra-args]
149 | "Returned OK but str was still NULL");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'with_filled_context_setup':
test/basictests.c:161:21: error: too many arguments for format [-Werror=format-extra-args]
161 | "Unable to intern 'one'");
| ^~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:163:21: error: too many arguments for format [-Werror=format-extra-args]
163 | "Unable to intern 'two'");
| ^~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:165:21: error: too many arguments for format [-Werror=format-extra-args]
165 | "Unable to intern 'three'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:167:21: error: too many arguments for format [-Werror=format-extra-args]
167 | "Unable to intern 'YAY'");
| ^~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:169:47: error: too many arguments for format [-Werror=format-extra-args]
169 | fail_unless(intern_one != intern_two, "'one' == 'two'");
| ^~~~~~~~~~~~~~~~
test/basictests.c:170:49: error: too many arguments for format [-Werror=format-extra-args]
170 | fail_unless(intern_one != intern_three, "'one' == 'three'");
| ^~~~~~~~~~~~~~~~~~
test/basictests.c:171:49: error: too many arguments for format [-Werror=format-extra-args]
171 | fail_unless(intern_two != intern_three, "'two' == 'three'");
| ^~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_interning_works_fn':
test/basictests.c:189:21: error: too many arguments for format [-Werror=format-extra-args]
189 | "Unable to re-intern 'one'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:192:21: error: too many arguments for format [-Werror=format-extra-args]
192 | "Internalising of the string failed");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_intern_substring_fn':
test/basictests.c:201:21: error: too many arguments for format [-Werror=format-extra-args]
201 | "Unable to intern 'hre'");
| ^~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:204:21: error: too many arguments for format [-Werror=format-extra-args]
204 | "Unable to re-intern 'hre' by substring");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:206:21: error: too many arguments for format [-Werror=format-extra-args]
206 | "'hre' != 'hre' -- wow!");
| ^~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_intern_substring_bad_offset_fn':
test/basictests.c:215:21: error: too many arguments for format [-Werror=format-extra-args]
215 | "Able to intern substring starting out of range");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_intern_substring_bad_size_fn':
test/basictests.c:224:21: error: too many arguments for format [-Werror=format-extra-args]
224 | "Able to intern substring ending out of range");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_ref_ok_fn':
test/basictests.c:231:21: error: too many arguments for format [-Werror=format-extra-args]
231 | "Oddly, reffing a string didn't return it");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_isequal_ok_fn':
test/basictests.c:253:21: error: too many arguments for format [-Werror=format-extra-args]
253 | "Failure comparing 'one' and 'two'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:255:21: error: too many arguments for format [-Werror=format-extra-args]
255 | "'one' == 'two' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_caseless_isequal_ok1_fn':
test/basictests.c:265:21: error: too many arguments for format [-Werror=format-extra-args]
265 | "Failure interning 'ONE'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:269:21: error: too many arguments for format [-Werror=format-extra-args]
269 | "'one' == 'ONE' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c:272:21: error: too many arguments for format [-Werror=format-extra-args]
272 | "Failure comparing 'one' and 'ONE' caselessly");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:274:21: error: too many arguments for format [-Werror=format-extra-args]
274 | "'one' !~= 'ONE' ?!");
| ^~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_caseless_isequal_ok2_fn':
test/basictests.c:284:21: error: too many arguments for format [-Werror=format-extra-args]
284 | "Failure interning 'yay'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:288:21: error: too many arguments for format [-Werror=format-extra-args]
288 | "'yay' == 'YAY' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c:291:21: error: too many arguments for format [-Werror=format-extra-args]
291 | "Failure comparing 'yay' and 'YAY' caselessly");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:293:21: error: too many arguments for format [-Werror=format-extra-args]
293 | "'yay' !~= 'YAY' ?!");
| ^~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_caseless_isequal_bad_fn':
test/basictests.c:302:21: error: too many arguments for format [-Werror=format-extra-args]
302 | "Failure comparing 'YAY' and 'one' caselessly");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:304:21: error: too many arguments for format [-Werror=format-extra-args]
304 | "'YAY' ~= 'one' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_extract_data_ok_fn':
test/basictests.c:313:21: error: too many arguments for format [-Werror=format-extra-args]
313 | "Extracting data ptr etc failed");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_is_nul_terminated_fn':
test/basictests.c:328:21: error: too many arguments for format [-Werror=format-extra-args]
328 | "Failure interning 'ONE'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:331:21: error: too many arguments for format [-Werror=format-extra-args]
331 | "Interned string isn't NUL terminated");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_substring_is_nul_terminated_fn':
test/basictests.c:341:21: error: too many arguments for format [-Werror=format-extra-args]
341 | "Failure interning 'ONE'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:344:21: error: too many arguments for format [-Werror=format-extra-args]
344 | "Failure interning substring 'O'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:347:21: error: too many arguments for format [-Werror=format-extra-args]
347 | "Interned substring isn't NUL terminated");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_tolower_ok1_fn':
test/basictests.c:358:21: error: too many arguments for format [-Werror=format-extra-args]
358 | "Failure interning 'ONE'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:361:38: error: too many arguments for format [-Werror=format-extra-args]
361 | fail_unless(result == false, "'one' == 'ONE' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c:363:37: error: too many arguments for format [-Werror=format-extra-args]
363 | fail_unless(result == true, "'one' != 'one' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_tolower_ok2_fn':
test/basictests.c:374:21: error: too many arguments for format [-Werror=format-extra-args]
374 | "Failure interning 'ONE'");
| ^~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:377:37: error: too many arguments for format [-Werror=format-extra-args]
377 | fail_unless(result == true, "'one' != 'ONE' (caseless) ?!");
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
test/basictests.c:380:38: error: too many arguments for format [-Werror=format-extra-args]
380 | fail_unless(result == false, "'one' == 'ONE' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c:382:37: error: too many arguments for format [-Werror=format-extra-args]
382 | fail_unless(result == true, "'one' != 'one' ?!");
| ^~~~~~~~~~~~~~~~~~~
test/basictests.c: In function 'test_lwc_string_iteration_fn':
test/basictests.c:398:35: error: too many arguments for format [-Werror=format-extra-args]
398 | fail_unless(counter == 4, "Incorrect string count");
| ^~~~~~~~~~~~~~~~~~~~~~~~
cc1: all warnings being treated as errors

0002886: Fails to build on i386 arch

Thu, 09 Jan 2025 06:38:20 +0000

When building libcss for i386 architecture, I get the following error:

COMPILE: test/parse2-auto.c
In file included from test/css21.c:11:
test/dump.h: In function 'dump_rule_media':
test/dump.h:134:45: error: format '%lx' expects argument of type 'long unsigned int', but argument 4 has type 'uint64_t' {aka 'long long unsigned int'} [-Werror=format=]
134 | ptr += sprintf(ptr, "| @media %s%03lx",
| ~~~~^
| |
| long unsigned int
| %03llx
135 | s->media->negate_type ? "not " : "",
136 | s->media->type);
| ~~~~~~~~~~~~~~
| |
| uint64_t {aka long long unsigned int}
cc1: all warnings being treated as errors
make[2]: *** [/usr/share/netsurf-buildsystem/makefiles/Makefile.top:649: build-i686-linux-gnu-i686-linux-gnu-release-lib-shared/test_css21.o] Error 1
make[2]: *** Waiting for unfinished jobs....
In file included from test/parse2-auto.c:12:
test/dump.h: In function 'dump_rule_media':
test/dump.h:134:45: error: format '%lx' expects argument of type 'long unsigned int', but argument 4 has type 'uint64_t' {aka 'long long unsigned int'} [-Werror=format=]
134 | ptr += sprintf(ptr, "| @media %s%03lx",
| ~~~~^
| |
| long unsigned int
| %03llx
135 | s->media->negate_type ? "not " : "",
136 | s->media->type);
| ~~~~~~~~~~~~~~
| |
| uint64_t {aka long long unsigned int}
cc1: all warnings being treated as errors

0002884: Save-To-Window doesn't dissapear after a "Save-All Shift-F3"

Sun, 05 Jan 2025 18:42:10 +0000

If you want so save the whole page of your aczually shown site, you could
press Shift-F3. A small window opens and you can drag the icon to a place,
where you want to save the page (--> e.g. "!bugs.netsurf.org").
In the earlier times ;-) this small window dissapeared after saving. Now
you have to explicitly close it.
If you save the page via clicking the menu-button of the mouse and site--> save all °f3, the same happens.

0002883: Does NetSurf support search in greek?

Mon, 02 Feb 2026 02:13:29 +0000

Hi. Does NetSurf (3.10) recognize greek? When I do a search in greek, it does not match any documents, and the key-words are displayed as question marks instead of letters.

0002882: Encountering

Sun, 22 Sep 2024 21:41:49 +0100

When NetSurf encounters an element, it seems to stop counting the height
of child elements and instead cuts it off, causing subsequent elements to overflow
from the parent, where as it is expected that the parent element should expand to fit the content, and indeed this is how it behaves in any other browser I have tested.

This specifically happens when the element is set to `display: block` and has an explicit height specified in CSS. Only both. Removing either stops the bug.

0002881: Crash in treeview after pressing left or right cursor keys

Wed, 11 Sep 2024 09:49:33 +0100

Cursor key navigation in a treeview without any previously selected node results in a crash.

Crash reproduced in both 3.11 and latest git update.

0002879: Can't Connect to Captive Portal Guest Network

Wed, 25 Jun 2025 12:19:11 +0100

Want to use Netsurf on my personal laptop at work. However, need to connect to guest network via a captive portal webpage to be able to click button accepting terms and conditions.

Netsurf tries to connect to page (I see the URL in the URL bar). However, it is not able to connect to the page.

First I see message on screen:
Privacy Error

A privacy error occurred while communicating with www.netsurf-browser.org
this may be a site configuration error or an attempt to steal private
information (passwords, messages or credit cards)

An unknown error occurred while processing the certificate.

The company I work for has a very high standard of security and use the guest network on some of my other laptops. So, I know network is safe.

I click on 'Proceed' button and then netsurf tries to load the logon page, but after a bit, I see:

[404] Resource Not Found.

Would greatly appreciate it if this can be fixed. On other computers, using browser like Firefox, I can connect to logon page OK.

Thank you.

Mark

0002878: Even/odd winding in SVG files rendered as NonZero

Fri, 21 Jun 2024 16:39:59 +0100

Holes in filled paths are rendered correctly only if the winding is NonZero - see the Drawfile in EvenOdd.zip and the exactly equivalent SVG file, which render correctly on other platforms, but differently from the Drawfile on NetSurf.

0002876: substack (and many other) web pages rendered too far right and will not scroll

Wed, 24 Dec 2025 20:32:30 +0000

All substack pages and many others render too far to the right so that text and graphics are partially out of view. Adjusting the window width does not help, and the horizontal scroll bar is always maxed out so you can't scroll the page into view.

0002875: Stack buffer overflow in idna_encode

Tue, 28 May 2024 14:54:41 +0100

This buffer overflow can be reached when NetSurf is handling "href" and "src" attributes in many different HTML elements, or when passing a URL on the command line.

In idna_encode, fqdn is created as a character buffer holding 256 bytes.
When called through nsurl_create and nsurl__create_from_section, the second argument, len, contains the correct value, but there is no limitations or checks on how long the host string can be.

This means that when strncpy is used to copy the provided host string to the fqdn buffer, it can be overflowed if the host string is longer than 256 bytes.
This causes a crash at minimum, and on builds without a stack canary it can be used to control the instruction pointer.

The HTML file shown below uses an offset of 312 bytes, followed by three f’s to set the last three bytes of the instruction pointer to \x66\x66\x66.
URL encoded bytes can also be used to change the instruction pointer to both lower and upper case ASCII. The number 3 is arbitrarily chosen to show a possible "relative jump" in a relocatable binary, but you can overwrite the entire instruction pointer or other things on the stack as well.

The offset 312 works for me on a 64-bit linux system with GCC, but might differ on other platforms.

0002874: Extreme resource use allowed in JavaScript

Tue, 28 May 2024 14:02:58 +0100

When NetSurf opens the HTML file shown below, it consumes 100% of a CPU core and at least 30 GB of memory, with no safeguards, meaning it will continue consuming memory until it is eventually done, or the system's OOM killer stops NetSurf or kills another process.

The root of this issue might be in the Duktape JavaScript engine itself, and be out of scope here?

However, it seems to me to be two-parted, both that the “normalize()”-function is very inefficient on this type of input, but also that neither NetSurf nor Duktape have any limits for duration or memory usage of a function like this.

The specific example shown here uses a complex series of left and right angle brackets to confuse the normalize-function, but I find it surprising that such a short HTML file can make the OOM killer take down my entire window manager.

0002873: Double free in box_normalise_table

Mon, 27 May 2024 09:38:42 +0100

Double free() caused by integer overflow when parsing a table in netsurf/content/handlers/html/box_normalise.c.

In box_normalise_table, col_info.spans is freed. If col_info.spans is already freed by box_normalise_table_row_group, this can cause a double free.
This can happen once `box_normalise_table_row_group` has called `box_normalise_table_row` which then calls `calculate_table_row`.
If cell_end_col is equal to 0xffffffff inside of calculate_table_row, the if-check on line 113 in netsurf/blob/master/content/handlers/html/box_normalise.c will evaluate to true, but passing `(cell_end_col + 1)` as the size argument to realloc, causing it to overflow to 0.
Calling realloc with size 0 is equivalent to a free, resulting in a double free triggered by the HTML file below.

The "magic constant" 4294967295 is (2^32)-1, filling an entire 32 bit integer.

0002872: Failure to handle http fetch from specific web site

Mon, 27 May 2024 13:54:28 +0100

Visiting https://stateofthenation.co/ results in error: Unrecognized or bad HTTP Content or Transfer-Encoding
Browsers on other platforms (Windows, iOS) render this site OK.
Iris also exhibits an issue with it, rendering a page of random text, so it looks like an encoding issue.
I tagged this as "always" reproducible but once, immediately after booting RISCOS, the page rendered correctly in NetSurf, but all subsequent attempts failed.

0002870: 0.9.2 tests fail to build on 32-bit system

Sat, 25 May 2024 12:14:42 +0100

When attempting to compile libcss tests for a i686-linux system, I get the following error:

```
In file included from test/css21.c:11:
test/dump.h: In function ‘dump_rule_media’:
test/dump.h:134:45: error: format ‘%lx’ expects argument of type ‘long unsigned int’, but argument 4 has type ‘uint64_t’ {aka ‘long long unsigned int’} [-Werror=format=]
134 | ptr += sprintf(ptr, "| @media %s%03lx",
| ~~~~^
| |
| long unsigned int
| %03llx
135 | s->media->negate_type ? "not " : "",
136 | s->media->type);
| ~~~~~~~~~~~~~~
| |
| uint64_t {aka long long unsigned int}
In file included from test/parse2-auto.c:12:
test/dump.h: In function ‘dump_rule_media’:
test/dump.h:134:45: error: format ‘%lx’ expects argument of type ‘long unsigned int’, but argument 4 has type ‘uint64_t’ {aka ‘long long unsigned int’} [-Werror=format=]
134 | ptr += sprintf(ptr, "| @media %s%03lx",
| ~~~~^
| |
| long unsigned int
| %03llx
135 | s->media->negate_type ? "not " : "",
136 | s->media->type);
| ~~~~~~~~~~~~~~
| |
| uint64_t {aka long long unsigned int}
COMPILE: test/select.c
LINK: build-i686-unknown-linux-gnu-i686-unknown-linux-gnu-release-lib-shared/test_lex
LINK: build-i686-unknown-linux-gnu-i686-unknown-linux-gnu-release-lib-shared/test_parse
LINK: build-i686-unknown-linux-gnu-i686-unknown-linux-gnu-release-lib-shared/test_csdetect
LINK: build-i686-unknown-linux-gnu-i686-unknown-linux-gnu-release-lib-shared/test_number
LINK: build-i686-unknown-linux-gnu-i686-unknown-linux-gnu-release-lib-shared/test_lex-auto
LINK: build-i686-unknown-linux-gnu-i686-unknown-linux-gnu-release-lib-shared/test_parse-auto
cc1: all warnings being treated as errors
```

The format specifier `%lx` does not match the `uint64_t` declaration on 32-bit systems, where `long int` is not 64 bits wide.

Anyhow, the following patch appears to fix the issue:

```
--- a/test/dump.h
+++ b/test/dump.h
@@ -131,7 +131,7 @@ void dump_rule_media(css_rule_media *s, char **buf, size_t *buflen)
char *ptr = *buf;
css_rule *rule;

- ptr += sprintf(ptr, "| @media %s%03lx",
+ ptr += sprintf(ptr, "| @media %s%03" PRIx64,
s->media->negate_type ? "not " : "",
s->media->type);
```

since the header is already in use.

0002869: hotlist issue with German territory

Sun, 26 May 2024 17:27:51 +0100

All only (?) with German territory module loaded.
If a hotlist is present, Netsurf 3.11 crash at start.
Without present hotlist, Netsurf works but if I try to add an URL to hotlist, Netsurf crash.
No known problems with 3.10.

0002868: Crash (segfault) on NULL style of nested box when trying to get computed height

Sun, 03 Mar 2024 08:11:49 +0000

Thread 1 "netsurf-gtk3" received signal SIGSEGV, Segmentation fault.
css_computed_height (style=0x0, length=length@entry=0x7fffffffac80, unit=unit@entry=0x7fffffffac84) at src/select/computed.c:614
614 return get_height(style, length, unit);
(gdb) bt
#0 css_computed_height (style=0x0, length=length@entry=0x7fffffffac80, unit=unit@entry=0x7fffffffac84) at src/select/computed.c:614
#1 0x00005555555bf21f in layout_find_dimensions (unit_len_ctx=unit_len_ctx@entry=0x55555692e180, available_width=available_width@entry=1300, box=box@entry=0x555556d78100, style=0x555556d77f30, width=width@entry=0x555556d78180, height=height@entry=0x555556d78184, max_width=0x555556db00c4, min_width=0x555556db00c0, max_height=0x555556db00cc,
min_height=0x555556db00c8, margin=0x555556d78198, padding=0x555556d781a8, border=0x555556d781b8, viewport_height=-1) at content/handlers/html/layout_internal.h:524
0000002 0x00005555555bf5ae in layout_flex_ctx__populate_item_data (flex=0x555556d77b20, available_width=1300, ctx=0x555556dba360) at content/handlers/html/layout_flex.c:355
#3 layout_flex (flex=flex@entry=0x555556d77b20, available_width=1300, content=content@entry=0x55555692df10) at content/handlers/html/layout_flex.c:1085
#4 0x00005555555b9cba in layout_block_context (block=block@entry=0x555556d59ac0, viewport_height=viewport_height@entry=-1, content=0x55555692df10, content@entry=0xfffffffffffffd90) at content/handlers/html/layout.c:3792
#5 0x00005555555bd402 in layout_float (b=b@entry=0x555556d59ac0, width=width@entry=2000, content=0xfffffffffffffd90, content@entry=0x55555692df10) at content/handlers/html/layout.c:2617
#6 0x00005555555ba78e in layout_line (next_box=, content=, has_text_children=, indent=, cont=, cy=199, cx=, y=, width=, first=0x555556d59ac0) at content/handlers/html/layout.c:2829
#7 layout_inline_container (content=, cy=199, cx=0, cont=, width=, inline_container=) at content/handlers/html/layout.c:3515
#8 layout_block_context (block=block@entry=0x5555568b2a90, viewport_height=viewport_height@entry=1980, content=content@entry=0x55555692df10) at content/handlers/html/layout.c:3829
#9 0x00005555555b9f4d in layout_block_context (block=block@entry=0x5555568b2000, viewport_height=viewport_height@entry=1980, content=content@entry=0x55555692df10) at content/handlers/html/layout.c:3796
#10 0x00005555555be599 in layout_document (content=content@entry=0x55555692df10, width=, width@entry=3818, height=height@entry=1980) at content/handlers/html/layout.c:5443
#11 0x00005555555ad5b3 in html_reformat (c=0x55555692df10, width=3818, height=1980) at content/handlers/html/html.c:1071
#12 0x0000555555584130 in content__reformat (c=0x55555692df10, background=background@entry=false, width=, height=) at content/content.c:344
#13 0x00005555555841f4 in content_reformat (h=, background=background@entry=false, width=, height=) at content/content.c:324
#14 0x00005555555ea6b9 in browser_window_content_ready (bw=0x5555563fa850) at desktop/browser_window.c:851
#15 browser_window_callback (c=, event=0x7fffffffb3d0, pw=0x5555563fa850) at desktop/browser_window.c:1500
#16 0x00005555555d2e54 in hlcache_content_callback (c=, msg=, data=, pw=) at content/hlcache.c:205
#17 0x0000555555583fd3 in content_broadcast (c=0x55555692df10, msg=CONTENT_MSG_READY, data=0x0) at content/content.c:762
#18 0x00005555555af007 in html_box_convert_done (c=0x55555692df10, success=) at content/handlers/html/html.c:276
#19 0x000055555559dbaf in convert_xml_to_box (ctx=0x555556c2f880) at content/handlers/html/box_construct.c:1300
#20 0x0000555555618025 in schedule_run () at frontends/gtk/schedule.c:148
#21 0x000055555560d5e8 in nsgtk_main () at frontends/gtk/gui.c:1072
#22 0x0000555555582eb1 in main (argc=, argv=) at frontends/gtk/gui.c:1231

0002867: Segmentation fault if form contains radio inputs without names

Sun, 25 Feb 2024 21:51:51 +0000

A radio button without a name causes a segmentation fault due to string comparison of null strings.