I'm playing around with rEFInd with logging turned on. I suspect when choosing certain boot items the system "shifts gears" before the logfile has been fully flushed to disk. With the ESP on a USB stick I've seen refind.log become corrupted along with the ESP's file allocation table on nearly every boot. On SATA the problem occurs with less frequency.
I don't see any discussion regarding this so I reckon the problem is limited to my system -- it's either ignoring or lying about fsync and there's not an actual rEFInd bug. Just wondering if there might be a workaround here -- perhaps I could compile rEFInd with a pause before firing up the boot entry?
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Can you temporarily rename your rEFInd efi file and drop the RefindPlus efi in instead (rename to match the original rEFInd efi name) to see whether this issue exists there as well?
I'm playing around with rEFInd with logging turned on. I suspect when choosing certain boot items the system "shifts gears" before the logfile has been fully flushed to disk. With the ESP on a USB stick I've seen refind.log become corrupted along with the ESP's file allocation table on nearly every boot. On SATA the problem occurs with less frequency.
I don't see any discussion regarding this so I reckon the problem is limited to my system -- it's either ignoring or lying about fsync and there's not an actual rEFInd bug. Just wondering if there might be a workaround here -- perhaps I could compile rEFInd with a pause before firing up the boot entry?
Can you temporarily rename your rEFInd efi file and drop the RefindPlus efi in instead (rename to match the original rEFInd efi name) to see whether this issue exists there as well?
https://github.com/dakanji/RefindPlus#installation
Last edit: dakanji 2023-06-18