Hello Minghua, minghua d: > The first error code 4294967268 is E4, or -28, -ENOSPC. Reading the
Exactly. It means ENOSPC and aufs cannot work correctly. > jjfs2.pdf from the web, it states: "If there is not enough dirty space > ... the kernel thread will sleep, and wirtes will fail with -ENOSPC > errors". Have you seen this before? Do you think I'm hitting this > situation? I think differently a little bit since the error happend around the XINO files. The aufs XINO files are hidden regular files and used internally. In your case, the size of the XINO files grows too much and got ENOSPC. It may be good to truncate the XINO files for you. While I don't remember the detail of aufs2.2 (which is unsupported now), the truncation might not be enabled by default for jffs2. Read the auf smanual and try one of these aufs mount options - trunc_xino - trunc_xino_v=%d:%d - trunc_xino=%s - itrunc_xino=%d - trunc_xib You can confirm the size and the consumed number of blocks via /debug/aufs/si_<id> dir if you mount debugfs. Or you can put the XINO files other that your first writable branch by specifying "xino=" mount option. > I patched the code to reduce the file name length by 1 so that it > builds. Pasted below is the patch. Hope this is not the cause to the > problem. Any comment? It doesn't seem to be related to the problem. But why do you need to decrement PATH_MAX? > By the way, aufs has been a very thoughtful project in both its design > and implementation. Greatly appreciate your work on it and hope > everything goes well with it! Thank you. J. R. Okajima ------------------------------------------------------------------------------ This SF.net email is sponsored by Windows: Build for Windows Store. http://p.sf.net/sfu/windows-dev2dev
