Automatic Release Builder
c0323e34a5
TerraSync: tweak warnings around checksum failures
...
Trying to reduce log spam when there is a server-side failure.
2020-10-29 14:00:22 +00:00
Automatic Release Builder
1ebb0787ad
Fix level of terrasync start/stop messages
2020-10-29 14:00:12 +00:00
Automatic Release Builder
5e37c2092a
Harden Repo::computeHashForPath
...
Check for some error cases in computeHashForPath, since this showed up
in some crash reports.
2020-10-29 14:00:01 +00:00
Automatic Release Builder
b3fc8deb5c
Exceptions: optional callback for exception throws.
...
This is to allow recording an error+stacktrace whenever an exception
is thrown, since that’s the point when the stack is interesting.
2020-10-29 13:59:44 +00:00
James Turner
f489232cdb
Fix for crash reported by Michael Danilov
...
In case uDNS returns a NULL txt pointer, don’t try to create a
std::string from it, since this will crash,
See: https://sourceforge.net/p/flightgear/codetickets/2398/
2020-10-25 18:59:37 +00:00
Scott Giese
f393a246ac
METAR: When Wind unit not specified, default to knots.
...
FLIGHTGEAR-F6 resolved
2020-10-23 23:26:28 -05:00
Scott Giese
d7a149e8cd
METAR Wind Sensor Failure.
...
FLIGHTGEAR-F9 resolved.
2020-10-23 22:12:51 -05:00
Erik Hofman
4a660e3827
callers already compensate for year-from-1900 and mont-start-from-0
2020-10-22 10:45:06 +02:00
Erik Hofman
ce187e614a
Fix the way struct tm works, year is since 1900 (so fix a year-2000 problem...) and month is 0-11 instead of 1-12. Return a string we constructed ourselves to prevent a possible buffer overrflow. Luckily the function isn't used in active code.
2020-10-22 10:20:37 +02:00
James Turner
83500dd413
CMake: use same 3rdParty logic as FlightGear
...
Avoids need to specify the 3rd-party dir, in the standard/fgmeta
layout of source directories, and accept more combinations of path,
when the user specifies a value.
2020-10-20 16:51:20 +01:00
Automatic Release Builder
2ad6e351ac
Update version file
2020-10-13 22:28:42 +01:00
Bertrand Coconnier
8c51d30ddb
Make CMake stop complaining about target "Udns::Udns" not found when rebuilding SimGear.
2020-10-03 17:58:11 +02:00
James Turner
670460a079
Set GCC RelwtihDebInfo opt flags
2020-09-29 17:40:25 +01:00
Scott Giese
cb4f3b1ca9
REVIEW: Memory Leaks
2020-09-24 22:24:12 -05:00
James Turner
4560095c36
Change props.cxx to throw std::exception
...
Throwing std::string is slightly unconventional, switch to throwing
exceptions inheriting std::exception.
2020-09-10 11:25:15 +01:00
Stuart Buchanan
b279384aef
Check correctly for failed BUILDING_LIST
...
Previous additional checks for a value BUILDING_LIST entry
used an enum incorrectly, resulting in all BUILDING_LIST entries
failing. Now fixed to check bad() / fail() which are bool.
2020-09-09 21:44:20 +01:00
Erik Hofman
01a190fea8
Fix a bug in the rotation_matrix function
2020-09-09 15:50:25 +02:00
James Turner
1341d1e1f2
Missed fix for logging in the test-suite
2020-09-08 12:34:37 +01:00
James Turner
051c920a99
Fix Linux compilation
...
Missing <cstring> header for strdup.
2020-09-08 08:15:41 +01:00
James Turner
0efcc14e91
Add MANDATORY_INFO log priority.
...
New log-level to avoid using ‘ALERT’ for expected mandatory messages,
since I want to do error collection on real alerts in a future commit.
As part of this, split logstream.hxx into separate header files, and
add a new virtual hook to LogCallback which takes the complete log-entry
2020-09-07 14:44:29 +01:00
James Turner
c7f013e7e5
Don’t wipe startup message on pausing logging.
2020-09-07 14:42:16 +01:00
James Turner
19ace8f0f9
Silence an alert message
2020-09-06 14:48:17 +01:00
James Turner
908496d43d
ASan: make copied files in log entries safe at init.
...
Ensure that if we copy file-names of log entries while startup
logging is active, we free them, but only once startup logging is
disabled, or on shutdown. This is needed to avoid crashes when
we use file-name copying for the Qt message handler.
2020-09-05 10:43:01 +01:00
James Turner
52c0f04e6e
Bindings: don’t cache the command pointer
...
Caching it complicates add/remove command logic, so making a simple
fix for now, which can be back-ported to 2020.2; ideally we would
cache the pointer but have an invalidation scheme, but that’s
considerably more work and risk.
Relates to Sentry crash:
https://sentry.io/organizations/flightgear/issues/1858764364
2020-09-04 10:51:32 +01:00
James Turner
f6656354b8
Refactor SGBuildingBin
...
Trying to trace a crash that was reported, but along the way fix a leak
of BuildingBins. Also avoid allocating some std::strings on the heap.
2020-09-03 20:35:50 +01:00
Erik Hofman
4928509646
Mipmapping requires a power-of-two destination buffer even if the source isn't. Neglecting this causes a massive bufferoverflow of the destination buffer. So for now we do not mipmap non-power-of-two textures and throw a warning. This affects at least the PC-9M and F-16, presumable many others. The effect will be a black texture when zoomed out and the proper texture when zooming in close enough.
2020-08-29 14:45:47 +02:00
Scott Giese
9ecb90edda
Clean up memory leak comment
2020-08-27 23:41:19 -05:00
Scott Giese
0d5552851b
Memory Leak Fixed - SGTexturedTriangleBin
...
Reclaimed 42,560 bytes in 280 blocks
2020-08-27 23:40:25 -05:00
James Turner
6f2b81b477
Fix for the fact we find_package(AAX) all the time.
...
Since we probe for AeonWave first, we set AAX_LIBRARY to NOT_FOUND,
which then upsets CMake if we try to use it. So add some temporary
ugly logic, until I have time to define a proper ALIAS target for the
audio backend.
2020-08-24 11:14:16 +01:00
Erik Hofman
af1b170059
Use consistent naming
2020-08-24 10:58:55 +02:00
Erik Hofman
f235a81519
Temporary fix for AeonWave support
2020-08-24 10:58:22 +02:00
Scott Giese
5b0f44f0db
Memory Leak Fix - matlib
...
Reclaimed 61,560 bytes in 405 blocks
2020-08-24 00:20:59 -05:00
Scott Giese
f14bcfb3c5
Memory Leak Fix - SGMaterial
...
Reclaimed 61,560 bytes in 405 blocks
2020-08-24 00:20:26 -05:00
Scott Giese
e1e70e8d33
Memory Leak Fix - BVHStaticGeometryBuilder
...
Was resolved at some point over the last two weeks.
Reclaimed 4,518,880 bytes in 141,215 blocks
2020-08-23 22:38:06 -05:00
Scott Giese
bd01e741ac
Memory Leak Fix - sg_path
...
Resolved via HTTPRespository fix
Reclaimed 272,181 bytes in 2,833 blocks
2020-08-23 22:22:18 -05:00
Scott Giese
423a0f65b0
Memory Leak Fix - HTTPRepository
...
Reclaimed 1,797,012 bytes in 7599 blocks
2020-08-23 22:15:28 -05:00
Scott Giese
b8e06453ea
Temporary fix: SimGear OpenAL dependency
...
FG configuration fails due to SG find_dependency(OpenAL) issue.
Skipping this dependency since FG itself will find_package(OpenAL).
2020-08-23 20:15:18 -05:00
James Turner
bab083d072
For static builds, ensure we find the sound library
...
Should fix Erik’s issue with AeonWave.
2020-08-23 18:10:49 +01:00
James Turner
2c4120974c
Follow rule of three for HTTPRepo:ChildInfo
2020-08-23 18:10:17 +01:00
James Turner
8b14e56f88
HTTPRepo: Fix ownership of HTTPDirectory
2020-08-23 18:08:49 +01:00
James Turner
3298260f11
GDAL is a public dependency :(
2020-08-23 11:27:09 +01:00
James Turner
2ed15eebfd
CMake fixes for Linux
2020-08-23 11:27:09 +01:00
Scott Giese
720a1fe7d2
REVIEW: Memory Leak - naTempSave()
...
1,024 bytes in 1 blocks are still reachable
2020-08-23 01:02:27 -05:00
Scott Giese
fc1178fb39
REVIEW: Memory Leak - ODGauge
...
1,176 bytes in 3 blocks are indirectly lost
2020-08-23 00:57:39 -05:00
Scott Giese
7ac400a162
REVIEW: Memory Leak - vgCreateContextSH()
...
1,624 bytes in 1 blocks are still reachable
2020-08-23 00:27:31 -05:00
Scott Giese
3550047bc8
REVIEW: Memory Leak - condition.cxx
...
2,880 bytes in 40 blocks are indirectly lost
632 bytes in 79 blocks are indirectly lost
2020-08-23 00:07:19 -05:00
Scott Giese
b7d2eea4b2
REVIEW: Memory Leak - props.cxx
...
2,028 (1,976 direct, 52 indirect) bytes in 13 blocks are definitely lost
1,963 bytes in 47 blocks are indirectly lost
1,544 bytes in 193 blocks are indirectly lost
662 bytes in 32 blocks are indirectly lost
152 bytes in 1 blocks are definitely lost
32 bytes in 1 blocks are indirectly lost
8 bytes in 1 blocks are indirectly lost
2020-08-22 23:52:36 -05:00
Scott Giese
69ad4a39e3
REVIEW: Memory Leak - nasal newBlock()
...
168 bytes in 7 blocks are still reachable
2020-08-22 22:42:09 -05:00
Scott Giese
4ccaa68c6a
REVIEW: Memory Leak - ODGauge
...
3,572 (136 direct, 3,436 indirect) bytes in 1 blocks are definitely lost
3,572 (8 direct, 3,564 indirect) bytes in 1 blocks are definitely lost
2,712 bytes in 3 blocks are indirectly lost
384 bytes in 3 blocks are indirectly lost
168 bytes in 3 blocks are indirectly lost
2020-08-22 22:38:33 -05:00
Scott Giese
fd6e4e7298
REVIEW: Memory Leak - nasal string.c; setlen()
...
3,953 bytes in 184 blocks are still reachable
2020-08-22 21:53:47 -05:00