Описание
FreeRDP is a free implementation of the Remote Desktop Protocol. Prior to 3.24.0, a size_t underflow in the IMA-ADPCM and MS-ADPCM audio decoders leads to heap-buffer-overflow write via the RDPSND audio channel. In libfreerdp/codec/dsp.c, the IMA-ADPCM and MS-ADPCM decoders subtract block header sizes from a size_t variable without checking for underflow. When nBlockAlign (received from the server) is set such that size % block_size == 0 triggers the header parsing at a point where size is smaller than the header (4 or 8 bytes), the subtraction wraps size to ~SIZE_MAX. The while (size > 0) loop then continues for an astronomical number of iterations. This vulnerability is fixed in 3.24.0.
Пакеты
| Пакет | Статус | Версия исправления | Релиз | Тип |
|---|---|---|---|---|
| freerdp3 | fixed | 3.24.0+dfsg-1 | package | |
| freerdp2 | removed | package |
Примечания
https://github.com/FreeRDP/FreeRDP/security/advisories/GHSA-85x9-4xxp-xhm5
Fixed by: https://github.com/FreeRDP/FreeRDP/commit/16df2300e1e3f5a51f68fb1626429e58b531b7c8 (3.24.0)
EPSS
Связанные уязвимости
FreeRDP is a free implementation of the Remote Desktop Protocol. Prior to 3.24.0, a size_t underflow in the IMA-ADPCM and MS-ADPCM audio decoders leads to heap-buffer-overflow write via the RDPSND audio channel. In libfreerdp/codec/dsp.c, the IMA-ADPCM and MS-ADPCM decoders subtract block header sizes from a size_t variable without checking for underflow. When nBlockAlign (received from the server) is set such that size % block_size == 0 triggers the header parsing at a point where size is smaller than the header (4 or 8 bytes), the subtraction wraps size to ~SIZE_MAX. The while (size > 0) loop then continues for an astronomical number of iterations. This vulnerability is fixed in 3.24.0.
FreeRDP is a free implementation of the Remote Desktop Protocol. Prior to 3.24.0, a size_t underflow in the IMA-ADPCM and MS-ADPCM audio decoders leads to heap-buffer-overflow write via the RDPSND audio channel. In libfreerdp/codec/dsp.c, the IMA-ADPCM and MS-ADPCM decoders subtract block header sizes from a size_t variable without checking for underflow. When nBlockAlign (received from the server) is set such that size % block_size == 0 triggers the header parsing at a point where size is smaller than the header (4 or 8 bytes), the subtraction wraps size to ~SIZE_MAX. The while (size > 0) loop then continues for an astronomical number of iterations. This vulnerability is fixed in 3.24.0.
EPSS