NishanthMenon changed the topic of #openocd to: this is the place to discuss all things OpenOCD | Logs: https://libera.irclog.whitequark.org/openocd/
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #openocd
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #openocd
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #openocd
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #openocd
borneoa___ has joined #openocd
_whitelogger_ has quit [Remote host closed the connection]
_whitelogger_ has joined #openocd
Hawk777 has joined #openocd
nerozero has joined #openocd
Hawk777 has quit [Quit: Leaving.]
jmi2k has joined #openocd
<jmi2k> Hello! I want to load an ELF file with OpenOCD but I want to skip sections like .bss and .heap (basically, those marked as NOLOAD in my linker script). Apparently, OpenOCD is clearing those areas, which is quite slow depending on the .heap size. Is there something I'm not understanding? Can it be done?
<dinkelhacker> Have you checked the program headers of your elf? Are these sections in a segment that doesn't have a LOAD type?
<jmi2k> I'm fairly sure it is in a NOLOAD segment, but I'll tell you for sure later
_whitelogger_ has quit [Remote host closed the connection]
<jmi2k> yeah, for some reason it ends up in the LOAD segment. Hold on, I will try to solve that
_whitelogger_ has joined #openocd
<dinkelhacker> I think the easiest way to test would be to quickly define a new segment in the MEMORY section and assign the sections with NOLOAD to that. I guess right now all your sections go into the same segemnt and if some of the are LOAD sections the whole segmement becomes LOAD. At least that would be my educated guess :)
<jmi2k> That's my current hypothesis, thanks for confirming it's a valid thing to consider haha
vampirefrog has quit [Ping timeout: 244 seconds]
<dinkelhacker> did it work?
<jmi2k> I'm trying to get a solution I'm happy with. Ideally, I'd want NOLOAD regions to follow LOAD regions in the same memory region.
<jmi2k> OTOH, I found that changing the definition I had for the heap space works well enough as a stopgap solution (as zeroing the heap memory area from OpenOCD is, by far, the biggest offender when it comes to load time)
vampirefrog has joined #openocd
_whitelogger_ has quit [Remote host closed the connection]
Guest60 has joined #openocd
Guest60 has quit [Client Quit]
_whitelogger_ has joined #openocd
nerozero has quit [Ping timeout: 255 seconds]
jmi2k has quit [Quit: Client closed]
jmi2k has joined #openocd
mrec has quit [Changing host]
mrec has joined #openocd
jmi2k has quit [Quit: Client closed]
_whitelogger_ has quit [Remote host closed the connection]
joconor has quit [Quit: ZNC 1.8.2 - https://znc.in]
joconor has joined #openocd
Hawk777 has joined #openocd
Hawk777 has quit [Quit: Leaving.]
WeaselSoup has quit [Quit: Q]
vampirefrog has quit [Quit: Leaving]
WeaselSoup has joined #openocd
vampirefrog has joined #openocd
vampirefrog has quit [Remote host closed the connection]
mrec has quit [Remote host closed the connection]