search for: sendfilev64

...lt;20070809035039.JDYJ10720.mta2.pacific.net.sg at mx3.pacific.net.sg>") BODY ("text" "html" ("charset" "Big5") NIL NIL "base64" 801726 10279)) A truss of the process reveals: write(1, " * 4 0 F E T C H (".., 36) = 36 sendfilev64(1, 1, 0xFFBFF2C0, 1, 0xFFBFF2B4) = 65536 close(10) = 0 close(8) = 0 open64("/var/qmail/maildirs/09/44/XXXXXX/Maildir/maildirsize", O_RDWR|O_APPEND) = 8 read(8, " 1 0 4 8 5 7 6 0 S\n 6 4".., 5120)...

...;s ok but when i attempt to read a message dovecot imap process crash. Dovecot crash even if i attempt to read one single email moved from my local inbox to remote imap INBOX maildir. Dovecot work fine on every folder created by maildrop. I made some debugging with truss and found that he die on sendfilev64 call: 5113: 0 7 O K S t a t u s c o m p l e t e d .\r\n 5113: poll(0x0008B450, 1, 204) = 0 5113: poll(0x0008B450, 1, 0) = 0 5113: poll(0x0008B450, 1, 999) = 1 5113: time()...

Hi, related to my previous email i found that setting "mail_save_crlf = no" instead of yes make dovecot working. However if enabled on solaris 8 it doesn't work hanging on sendfilev64() call. I noticed that maildir file names are still different from maildrop and dovecot but now works. However moving bigger mailbox from local storage to remote imap server i receive "Disconnect for inactivity while waiting for command data" error triggered by client_input_timeout func...

Hi Timo and other dovecot fans, why, when I'm trying to read certain "Drafts" emails that I just created, or edited, do the dovecot children die (or so I think). In mutt, the behaviour is that I "c =Drafts" then select the email and I get "connection closed to squeaky.rubberduck.com". In Mail.app - she just loops, trying to read those headers again and again

...... no checking whether to support ACLs... auto configure: checking whether ACL support is available: configure: Using solaris ACLs checking whether to support asynchronous io... no checking whether to check to support sendfile... yes checking for sendfilev in -lsendfile... yes checking for solaris sendfilev64 support... yes checking for solaris sendfilev support... yes checking for Linux readahead... no checking for readahead declaration... no checking for posix_fadvise... no checking for splice declaration... no checking for a broken readlink syscall... no checking whether to build winbind... yes check...