+- PINE64 (https://forum.pine64.org)
+-- Forum: PineTab (https://forum.pine64.org/forumdisplay.php?fid=140)
+--- Forum: General Discussion on PineTab (https://forum.pine64.org/forumdisplay.php?fid=141)
+--- Thread: Possible solution to boot and/or eMMC failures (/showthread.php?tid=19772)
Possible solution to boot and/or eMMC failures - sanctimonious_spoon - 04-01-2025
After attempting to flash an image to the eMMC, my Pinetab 2 would not boot from the eMMC, and when I booted from the SD card, I saw an error similar to the following
Code:
mmc0: error -110 whilst initialising MMC cardI tried flashing the eMMC with the factory image multiple times, but none of the attempts succeeded.
Based on my experience with Xilinx SoCs, I suspected an issue with the SD Host Controller (SDHC). I poked at some of the internal registers on the SDHC in U-Boot, but while I noticed some possible patterns, I couldn't pinpoint a definitive issue or find any magical reset sequence that fixed the issue. I was able to observe that some of the low-level SD commands were timing out due to lack of response. Off a hunch, I removed the back cover and disconnected the battery cable form the main board (I also disconnected the other connectors so I could examine the main board, but given that both the SDHC and the eMMC are located on the main board, I wouldn't expect the removal of any of these to have an impact). When I reinserted the cable and booted the Pinetab 2 I was able to write an image to the eMMC and subsequently boot from the eMMC. My theory is that the SD/MMC/eMMC state machine in the SDHC was somehow corrupted, leaving it unable to enter a state where it could respond to commands; I suspect that fully de-energizing and re-energizing the circuit caused it to go into reset, getting it back on an operational path. It's possible the issue is with the eMMC chip instead of the SDHC, but my knowledge of SD/MMC standards and implementation is limited and I was unable to find substantial documentation on the eMMC chip. As a caveat, I have not had the time to attempt to reproduce the issue or my solution, so it's possible my observation and solution were subject to confounding factors I didn't test for. (Time aside, I damaged the microSD card reader while attempting to reseat the back cover, so I would need to get USB booting to work in order to be able to effectively run tests.)
Hopefully, this is helpful to someone.