[Live-devel] Heap-user-after-free in live.2023.06.20

2023-06-23 Thread Meng Ruijie
Hello, There may be one heap-use-after-free while calling RTPInterface::sendDataOverTCP. The following is the bug report from the ASAN: ==17==ERROR: AddressSanitizer: heap-use-after-free on address 0x62e000f3a0e8 at pc 0x005d5ec8 bp 0x735fc420 sp 0x735fc418 READ of size 1 at 0

***UNCHECKED*** [Live-devel] stack-use-after-return while handling SETUP

2023-06-13 Thread Meng Ruijie
Here is one bug about stack-use-after-return in live.2023.05.10 running in Ubuntu 20.04. It happens while handling the SETUP client request. You can reproduce this bug as the README in the attachment. It is one flaky bug as we observed. The following is the ASAN bug report:

***UNCHECKED*** [Live-devel] heap-use-after-free while handling PLAY

2023-06-13 Thread Meng Ruijie
Hi, We found one heap-use-after-free in live.2023.05.10 running in Ubuntu 20.04. It happens while handling the PLAY client request by calling the function `MultiFramedRTPSink::continuePlaying`. We attached the relevant files to reproduce this bug and hope it could help you fix it. You can repr

***UNCHECKED*** [Live-devel] heap-use-after-free in continueParsing

2023-06-13 Thread Meng Ruijie
Hi, We found one heap-use-after-free bug in live.2023.05.10 while running in Ubuntu:20.04. This bug happens while calling the function `MatroskaFileParser::continueParsing`. You can reproduce this bug as the README in the attachment as follows: 1. build the docker image: docker build . -t u

[Live-devel] Invalid read in HandleCmd_DESCRIBE

2023-06-08 Thread Meng Ruijie
Hi, We found one memory issue about an invalid read in the function handleCmd_DECRIBE in live.2023.05.10. Here is the bug report from the Valgrind: --- ==1744== Invalid read of size 1 ==1744==at 0x483EF46: strlen (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so) ==1

[Live-devel] Memory leak in handleCmd_PLAY

2023-06-08 Thread Meng Ruijie
Hi, We found one memory leak in the function handleCmd_PLAY in live.2023.05.10. Here is the bug report from the valgrind: --- ==1755== Invalid read of size 1 ==1755==at 0x483EF46: strlen (in /usr/lib/x86_64-linux-gnu/valgrind/vgpreload_memcheck-amd64-linux.so) ==1755==by 0x4E2FD14: __v