Closed arvindpereira closed 7 years ago
Thanks for sending this over, Arvind, this looks like a good implementation. The thing that's holding me up is that I would like to make it possible to build on F411 as well, just without Lepton 3 support. I haven't had any time to work on this, but hopefully will soon.
Totally understand... I didn't have cycles to keep it backward compatible with the 411. I mean it is backward compatible with the 411 but only for Y16. That's the only thing I cared about. There is a flag to build it for the 411 although 412 is the default...
On Wed, May 24, 2017, 9:29 AM Kurt Kiefer notifications@github.com wrote:
Thanks for sending this over, Arvind, this looks like a good implementation. The thing that's holding me up is that I would like to make it possible to build on F411 as well, just without Lepton 3 support. I haven't had any time to work on this, but hopefully will soon.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/groupgets/purethermal1-firmware/pull/11#issuecomment-303778402, or mute the thread https://github.com/notifications/unsubscribe-auth/AEdbYP2C4d_fr28Bw5DFOVe3RRzpw_rfks5r9FrmgaJpZM4NMw7- .
Thanks for the V3 support! To build this project in eclipse you need to change the '.cproject' file to point the linker script to ../STM32F412CGUx_FLASH.ld file instead of the ../STM32F11*.ld file, otherwise the build process will fail due to lack of RAM. If you search for '.ld' within this file the change is very quick.
You're welcome. Well, I only used vim and the command line when working on it, so didn't bother with the.cproject file. You can supply the appropriate flags at build time along with make so it does the right thing. That being said, if you have a good way of being able to choose v2 vs v3 via eclipse so it builds the correct variant, please pull request to my branch.
On Jun 1, 2017 8:12 AM, "MatthewDeKoning" notifications@github.com wrote:
Thanks for the V3 support! To build this project in eclipse you need to change the '.cproject' file to point the linker script to ../STM32F412CGUx_FLASH.ld file instead of the ../STM32F11*.ld file, otherwise the build process will fail due to lack of RAM. If you search for '.ld' within this file the change is very quick.
— You are receiving this because you authored the thread. Reply to this email directly, view it on GitHub https://github.com/groupgets/purethermal1-firmware/pull/11#issuecomment-305524500, or mute the thread https://github.com/notifications/unsubscribe-auth/AEdbYHWBX8btLSuoGSdpbr8XCzjVosbKks5r_tTsgaJpZM4NMw7- .
I went ahead and merged this into Lepton3 here for further work. Thanks again for kicking this off.
Another quick question: how does your computer recognize the PureThermal board after flashing it with the Lepton3 binary file? I cannot find it as a device on windows for the life of me, but being on windows could be contributing to the problem. Maybe on Linux it can be accessed through /dev/video0 still?
…t backward compatible with Lepton 2. One should not see the loss of sync that usually happens with the current HEAD of PT-1 code for Lepton 3. A big caveat to being able to use this is that one needs to have the newer STM on the PT-1 i.e. the STM32F412CGU as opposed to the STM32F411CEU which was used on older STMs. A big change was modifying the linker script to ensure that we were utilizing the RAM on the F412CGU which allows our buffering between SPI and UVC transfers to happen glitch-free.