Hi, We are using FUSE for our userspace clustered Filesystem, and recently we noticed READ corruptions when we enable SAMBA AIO to access the file system. SAMBA version: 4.3.13 Kernel Version: 4.9.30 To troubleshoot this further we captured/Instrumented the CRC32 checksums in SAMBA just after the PREAD and against our native clustered Filesystem and at Kernel VFS pread64. With the simple replay program which does reading of the file where it returned the bad/corrupted read buffer against the offset and requested the length of data showed fuse kernel is printing the bad CRC32 against few couple reads. So, I am here to seek some help, on how to proceed to further debug the issue in fuse kernel.
