Just to update and close this conversation with a better understanding of my problem and providing solution:
I basically had 2 problems: 1. When calling *smart_initialize() *and *nx_mount()* , I had "no low level format" error. To fix that, I needed to force format using the following command: *mksmartfs -f /dev/smartn* where n is the minor number used in the *smart_initialize()* function. Then, I reset the board and mount was ok. Note: if you don't force (use -f) it won't work. 2. Some of the intermittent I/O errors I had that I was not able to describe, because it happened to some files but not to others, was a very simple isse: it was related to the filename's size. Which default is 16. Some configs you should select/configure when using SMART FS: CONFIG_FS_SMARTFS=y CONFIG_MTD_SMART=y CONFIG_NAME_MAX=64 CONFIG_SMARTFS_MAXNAMLEN=64 Em seg., 9 de ago. de 2021 às 10:38, Sara da Cunha Monteiro de Souza < saramonteirosouz...@gmail.com> escreveu: > Hi Ken, > > Thank you. I'll appreciate to know the results of your tests. > I am also performing more tests than this one and I am experiencing issues > that I still > don't know if it is related to SMART FS, or some NuttX module or even the > chip I am using. > The issues are intermittent, what makes it harder to identify the cause. > > Em sex., 6 de ago. de 2021 às 15:21, Ken Pettit <petti...@gmail.com> > escreveu: > >> Hi Sara, >> >> I'm not sure SmartFS was ever tested with this particular sequence where >> the volume is erase and then re-mounted to a different location. If I >> get a chance this weekend, I will take a look at it since you have a >> very specific test sequence. >> >> Ken >> >> On 8/5/21 3:48 PM, Sara da Cunha Monteiro de Souza wrote: >> > Hi all, >> > I am facing some issues when enabling SMART FS and I'd like to know if >> > this is an expected behavior or maybe a bug on NuttX. >> > >> > First issue: >> > - If I use the following sequence: >> > 1. smart_initialize() >> > 2. nx_mount() >> > nx_mount throws errors like: sector not allocated and no low level >> format. >> > >> > To fix it, all I need to do is to format it by calling mksmartfs >> > /dev/smart0 and I can reset and have my mount point correctly mounted. >> > >> > Second issue: >> > >> > If I run flash_eraseall just like the following sequence it will throw >> an >> > I/O error when writing to a file in the recent created mount point. >> > >> > mksmartfs /dev/smart0 mount -t smartfs /dev/smart0 /temp echo "TESTE" > >> > /temp/teste cat /temp/teste umount /temp flash_eraseall /dev/smart0 >> > mksmartfs /dev/smart0 mount -t smartfs /dev/smart0 /another_temp >> > echo "TESTE" > /another_temp/teste >> > >> > P.S: >> > The error happened on ESP32C3 DevKit and also on simulator. >> > >> > To reproduce the error, you basically need to enable the following >> configs: >> > CONFIG_FS_SMARTFS >> > CONFIG_MTD_SMART >> > CONFIG_SYSTEM_FLASH_ERASEALL >> > >> > Thanks, Sara. >> > >> >>
