From fe3f8e3e172818741f32f5ea0bba27f77ca56a1a Mon Sep 17 00:00:00 2001 From: "Suren A. Chilingaryan" Date: Fri, 14 Aug 2015 03:53:15 +0200 Subject: Add another workaround to handle cameras stuck in busy (disabled and untested) --- private.h | 1 + 1 file changed, 1 insertion(+) (limited to 'private.h') diff --git a/private.h b/private.h index 0e16044..38d6dd0 100644 --- a/private.h +++ b/private.h @@ -22,6 +22,7 @@ #define IPECAMERA_BUG_MULTIFRAME_PACKETS //**< This is by design, start of packet comes directly after the end of last one in streaming mode */ #define IPECAMERA_BUG_MULTIFRAME_HEADERS //**< UFO Camera operates with 32-byte entities, but some times there is 16-byte padding before the data which may result in spliting the header between 2 DMA packets. We still need to define a minimal number of bytes which are always in the same DMA packet (CMOSIS_ENTITY_SIZE) */ #define IPECAMERA_BUG_REPEATING_DATA //**< 16 bytes repeated at frame offset 4096, the problem start/stop happenning on board restart */ +//#define IPECAMERA_BUG_STUCKED_BUSY //**< DMA may stuck in busy. According to Michele, the work-around is to disable triggering and read all data out. Not checked */ //#define IPECAMERA_BUG_INCOMPLETE_PACKETS //**< Support incomplete packets, i.e. check for frame magic even if full frame size is not reached yet (slow) */ //#define IPECAMERA_ANNOUNCE_READY //**< Announce new event only after the reconstruction is done */ //#define IPECAMERA_CLEAN_ON_START //**< Read all the data from DMA before starting of recording */ -- cgit v1.2.3