<div class="gmail_quote">On Mon, Feb 22, 2010 at 10:53 AM, Javier Henderson <span dir="ltr"><<a href="mailto:javier@kjsl.org">javier@kjsl.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
(Client) Protocol versions: remote=29, negotiated=29<br>
Invalid flist flag: 1004<br></blockquote></div><div><br></div>I looked at both 2.6.8 and 2.6.9 last week and didn't see any way for the old code to construct a 0x1004 value for the file list flags, but sadly I neglected to answer you back. The older code doesn't even have an 0x1000 bit (1<<12), so that makes me think that some kind of corruption is occurring in the transmitted bytes. If you want to help diagnose the issue, add some debug code to the old 2.6.8 software to make it log a flag value of 0x1004. Then we would know if 0x1004 was indeed being transmitted by the code. From your usage, it appears that the 2.6.8 software is an rsync daemon, so lets have it write to the daemon log (edit flist.c):<div>
<br></div><div><div> if (protocol_version >= 28) {</div><div> if (!flags && !S_ISDIR(mode))</div><div> flags |= XMIT_TOP_DIR;</div><div> if ((flags & 0xFF00) || !flags) {</div>
<div> flags |= XMIT_EXTENDED_FLAGS;</div><div> if (flags == 0x1004) rprintf(FLOG, "Sending flags = %x!!\n", flags); /* <-- Add this line */</div><div> write_byte(f, flags);</div>
<div> write_byte(f, flags >> 8);</div><div> } else</div><div> write_byte(f, flags);</div><div> } else {</div><div><br></div>If you find that 0x1004 is indeed being sent, we'll have to figure out if there is any way (other than memory corruption) for that flag to be sent.</div>
<div><br>..wayne..<br>
</div>