KNOWNBUGS revision 182352
1101099Srwatson 2166905Srwatson 3145412Strhodes K N O W N B U G S I N S E N D M A I L 4171253Srwatson 5172930Srwatson 6101099SrwatsonThe following are bugs or deficiencies in sendmail that we are aware of 7101099Srwatsonbut which have not been fixed in the current release. You probably 8101099Srwatsonwant to get the most up to date version of this from ftp.sendmail.org 9145412Strhodesin /pub/sendmail/KNOWNBUGS. For descriptions of bugs that have been 10101099Srwatsonfixed, see the file RELEASE_NOTES (in the root directory of the sendmail 11106393Srwatsondistribution). 12106393Srwatson 13106393SrwatsonThis list is not guaranteed to be complete. 14106393Srwatson 15101099Srwatson* Delivery to programs that generate too much output may cause problems 16172930Srwatson 17172930Srwatson If e-mail is delivered to a program which generates too much 18172930Srwatson output, then sendmail may issue an error: 19101099Srwatson 20101099Srwatson timeout waiting for input from local during Draining Input 21101099Srwatson 22101099Srwatson Make sure that the program does not generate output beyond a 23101099Srwatson status message (corresponding to the exit status). This may 24101099Srwatson require a wrapper around the actual program to redirect output 25101099Srwatson to /dev/null. 26101099Srwatson 27101099Srwatson Such a problem has been reported for bulk_mailer. 28101099Srwatson 29101099Srwatson* Null bytes are not handled properly in headers. 30101099Srwatson 31101099Srwatson Sendmail should handle full binary data. As it stands, it handles 32101099Srwatson all values in the body, but not 0x00 in the header. Changing 33101099Srwatson this would require a major restructuring of the code -- for 34101099Srwatson example, almost no C library support could be used to handle 35101099Srwatson strings. 36101099Srwatson 37101099Srwatson* Header checks are not called if header value is too long or empty. 38101099Srwatson 39101099Srwatson If the value of a header is longer than 1250 (MAXNAME + MAXATOM - 6) 40101099Srwatson characters or it contains a single word longer than 256 (MAXNAME) 41101099Srwatson characters then no header check is done even if one is configured for 42136774Srwatson the header. 43101099Srwatson 44101099Srwatson* Header lines which are too long will be split incorrectly. 45171253Srwatson 46171253Srwatson Header lines which are longer than 2045 characters will be split 47171253Srwatson but some characters might be lost. Fix: obey RFC (2)822 and do not 48101099Srwatson send lines that are longer than 1000 characters. 49101099Srwatson 50101099Srwatson* milter communication fails if a single header is larger than 64K. 51101099Srwatson 52101099Srwatson If a single header is larger than 64KB (which is not possible in the 53157986Sdwmalone default configuration) then it cannot be transferred in one block to 54145412Strhodes libmilter and hence the communication fails. This can be avoided by 55101099Srwatson increasing the constant MILTER_CHUNK_SIZE in 56166905Srwatson include/libmilter/mfdef.h and recompiling sendmail, libmilter, and 57101099Srwatson all (statically linked) milters (or by using an undocumented compile 58145412Strhodes time option: _FFR_MAXDATASIZE; you have to read the source code in 59170689Srwatson order to use this properly). 60101099Srwatson 61101099Srwatson* Sender addresses whose domain part cause a temporary A record lookup 62101099Srwatson failure but have a valid MX record will be temporarily rejected in 63134132Strhodes the default configuration. Solution: fix the DNS at the sender side. 64101099Srwatson If that's not easy to achieve, possible workarounds are: 65165469Srwatson - add an entry to the access map: 66101099Srwatson dom.ain OK 67101099Srwatson - (only for advanced users) replace 68145412Strhodes 69145412Strhodes# Resolve map (to check if a host exists in check_mail) 70101099SrwatsonKresolve host -a<OKR> -T<TEMP> 71101099Srwatson 72101099Srwatson with 73101099Srwatson 74101099Srwatson# Resolve map (to check if a host exists in check_mail) 75101099SrwatsonKcanon host -a<OKR> -T<TEMP> 76101099SrwatsonKdnsmx dns -R MX -a<OKR> -T<TEMP> 77101099SrwatsonKresolve sequence dnsmx canon 78101099Srwatson 79101099Srwatson 80101099Srwatson* Duplicate error messages. 81101099Srwatson 82101099Srwatson Sometimes identical, duplicate error messages can be generated. As 83101099Srwatson near as I can tell, this is rare and relatively innocuous. 84101099Srwatson 85101099Srwatson* Misleading error messages. 86157986Sdwmalone 87101099Srwatson If an illegal address is specified on the command line together 88101099Srwatson with at least one valid address and PostmasterCopy is set, the 89101099Srwatson DSN does not contain the illegal address, but only the valid 90101099Srwatson address(es). 91101099Srwatson 92157986Sdwmalone* \231 considered harmful. 93157986Sdwmalone 94101099Srwatson Header addresses that have the \231 character (and possibly others 95134132Strhodes in the range \201 - \237) behave in odd and usually unexpected ways. 96171253Srwatson 97171253Srwatson* accept() problem on SVR4. 98134132Strhodes 99134132Strhodes Apparently, the sendmail daemon loop (doing accept()s on the network) 100134132Strhodes can get into a weird state on SVR4; it starts logging ``SYSERR: 101134132Strhodes getrequests: accept: Protocol Error''. The workaround is to kill 102134132Strhodes and restart the sendmail daemon. We don't have an SVR4 system at 103134132Strhodes Berkeley that carries more than token mail load, so I can't validate 104171253Srwatson this. It is likely to be a glitch in the sockets emulation, since 105171253Srwatson "Protocol Error" is not possible error code with Berkeley TCP/IP. 106171253Srwatson 107134131Strhodes I've also had someone report the message ``sendmail: accept: 108101099Srwatson SIOCGPGRP failed errno 22'' on an SVR4 system. This message is 109134131Strhodes not in the sendmail source code, so I assume it is also a bug 110134131Strhodes in the sockets emulation. (Errno 22 is EINVAL "Invalid Argument" 111171253Srwatson on all the systems I have available, including Solaris 2.x.) 112171253Srwatson Apparently, this problem is due to linking -lc before -lsocket; 113134131Strhodes if you are having this problem, check your Makefile. 114134131Strhodes 115101099Srwatson* accept() problem on Linux. 116101099Srwatson 117101099Srwatson The accept() in sendmail daemon loop can return ETIMEDOUT. An 118157986Sdwmalone error is reported to syslog: 119101099Srwatson 120157986Sdwmalone Jun 9 17:14:12 hostname sendmail[207]: NOQUEUE: SYSERR(root): 121101099Srwatson getrequests: accept: Connection timed out 122157986Sdwmalone 123157986Sdwmalone "Connection timed out" is not documented as a valid return from 124157986Sdwmalone accept(2) and this was believed to be a bug in the Linux kernel. 125157986Sdwmalone Later information from the Linux kernel group states that Linux 126157986Sdwmalone 2.0 kernels follow RFC1122 while sendmail follows the original BSD 127157986Sdwmalone (now POSIX 1003.1g draft) specification. The 2.1.X and later kernels 128157986Sdwmalone will follow the POSIX draft. 129136739Srwatson 130101099Srwatson* Excessive mailing list nesting can run out of file descriptors. 131101099Srwatson 132101099Srwatson If you have a mailing list that includes lots of other mailing 133101099Srwatson lists, each of which has a separate owner, you can run out of 134101099Srwatson file descriptors. Each mailing list with a separate owner uses 135101099Srwatson one open file descriptor (prior to 8.6.6 it was three open 136101099Srwatson file descriptors per list). This is particularly egregious if 137101099Srwatson you have your connection cache set to be large. 138101099Srwatson 139101099Srwatson* Connection caching breaks if you pass the port number as an argument. 140101099Srwatson 141145412Strhodes If you have a definition such as: 142101099Srwatson 143101099Srwatson Mport, P=[IPC], F=kmDFMuX, S=11/31, R=21, 144101099Srwatson M=2100000, T=DNS/RFC822/SMTP, 145101099Srwatson A=IPC [127.0.0.1] $h 146101099Srwatson 147154386Scsjp (i.e., where $h is the port number instead of the host name) the 148101099Srwatson connection caching code will break because it won't notice that 149101099Srwatson two messages addressed to different ports should use different 150145412Strhodes connections. 151145412Strhodes 152145412Strhodes* ESMTP SIZE underestimates the size of a message 153101099Srwatson 154101099Srwatson Sendmail makes no allowance for headers that it adds, nor does it 155145412Strhodes account for the SMTP on-the-wire \r\n expansion. It probably doesn't 156145412Strhodes allow for 8->7 bit MIME conversions either. 157101099Srwatson 158101099Srwatson* Client ignores SIZE parameter. 159145412Strhodes 160145412Strhodes When sendmail acts as client and the server specifies a limit 161145412Strhodes for the mail size, sendmail will ignore this and try to send the 162145412Strhodes mail anyway. The server will usually reject the MAIL command 163145412Strhodes which specifies the size of the message and hence this problem 164101099Srwatson is not significant. 165145412Strhodes 166145412Strhodes* Paths to programs being executed and the mode of program files are 167145412Strhodes not checked. Essentially, the RunProgramInUnsafeDirPath and 168145412Strhodes RunWritableProgram bits in the DontBlameSendmail option are always 169145412Strhodes set. This is not a problem if your system is well managed (that is, 170145412Strhodes if binaries and system directories are mode 755 instead of something 171145412Strhodes foolish like 777). 172145412Strhodes 173145412Strhodes* 8-bit data in GECOS field 174145412Strhodes 175145412Strhodes If the GECOS (personal name) information in the passwd file contains 176145412Strhodes 8-bit characters, those characters can be included in the message 177145412Strhodes header, which can cause problems when sending SMTP to hosts that 178145412Strhodes only accept 7-bit characters. 179145412Strhodes 180145412Strhodes* 8->7 bit MIME conversion 181101099Srwatson 182101099Srwatson When sendmail is doing 8->7 bit MIME conversions, and the message 183145412Strhodes contains certain MIME body types that cannot be converted to 7-bit, 184101099Srwatson sendmail will pass the message as 8-bit. 185101099Srwatson 186101099Srwatson* 7->8 bit MIME conversion 187145412Strhodes 188145412Strhodes If a message that is encoded as 7-bit MIME is converted to 8-bit and 189145412Strhodes that message when decoded is illegal (e.g., because of long lines or 190101099Srwatson illegal characters), sendmail can produce an illegal message. 191171253Srwatson 192101099Srwatson* MIME encoded full name phrases in the From: header 193101099Srwatson 194145412Strhodes If a full name phrase includes characters from MustQuoteChars, sendmail 195145412Strhodes will quote the entire full name phrase. If MustQuoteChars includes 196145412Strhodes characters which are not special characters according to STD 11 (RFC 197145412Strhodes 822), this quotation can interfere with MIME encoded full name phrases. 198148482Strhodes By default, sendmail includes the single quote character (') in 199145412Strhodes MustQuoteChars even though it is not listed as a special character in 200148482Strhodes STD 11. 201101099Srwatson 202101099Srwatson* bestmx map with -z flag truncates the list of MX hosts 203171253Srwatson 204171253Srwatson A bestmx map configured with the -z flag will truncate the list 205101099Srwatson of MX hosts. This prevents creation of strings which are too 206101099Srwatson long for ruleset parsing. This can have an adverse effect on the 207101099Srwatson relay_based_on_MX feature. 208101099Srwatson 209101099Srwatson* Saving to ~sender/dead.letter fails if su'ed to root 210145412Strhodes 211101099Srwatson If ErrorMode is set to print and an error in sending mail occurs, 212101099Srwatson the normal action is to print a message to the screen and append 213101099Srwatson the message to a dead.letter file in the sender's home directory. 214101099Srwatson In the case where the sender is using su to act as root, the file 215101099Srwatson safety checks prevent sendmail from saving the dead.letter file 216101099Srwatson because the sender's uid and the current real uid do not match. 217145412Strhodes 218101099Srwatson* Berkeley DB 2.X race condition with fcntl() locking 219101099Srwatson 220101099Srwatson There is a race condition for Berkeley DB 2.X databases on 221101099Srwatson operating systems which use fcntl() style locking, such as 222157986Sdwmalone Solaris. Sendmail locks the map before calling db_open() to 223101099Srwatson prevent others from modifying the map while it is being opened. 224101099Srwatson Unfortunately, Berkeley DB opens the map, closes it, and then 225157986Sdwmalone reopens it. fcntl() locking drops the lock when any file 226101099Srwatson descriptor pointing to the file is closed, even if it is a 227101099Srwatson different file descriptor than the one used to initially lock 228101099Srwatson the file. As a result there is a possibility that entries in a 229101099Srwatson map might not be found during a map rebuild. As a workaround, 230145412Strhodes you can use makemap to build a map with a new name and then 231157986Sdwmalone "mv" the new db file to replace the old one. 232157986Sdwmalone 233157986Sdwmalone Sleepycat Software has added code to avoid this race condition to 234157986Sdwmalone Berkeley DB versions after 2.7.5. 235157986Sdwmalone 236157986Sdwmalone* File open timeouts not available on hard mounted NFS file systems 237157986Sdwmalone 238157986Sdwmalone Since SIGALRM does not interrupt an RPC call for hard mounted 239101099Srwatson NFS file systems, it is impossible to implement a timeout on a file 240101099Srwatson open operation. Therefore, while the NFS server is not responding, 241101099Srwatson attempts to open a file on that server will hang. Systems with 242101099Srwatson local mail delivery and NFS hard mounted home directories should be 243101099Srwatson avoided, as attempts to open the forward files could hang. 244157986Sdwmalone 245157986Sdwmalone* Race condition for delivery to set-user-ID files 246157986Sdwmalone 247157986Sdwmalone Sendmail will deliver to a fail if the file is owned by the DefaultUser 248157986Sdwmalone or has the set-user-ID bit set. Unfortunately, some systems clear that bit 249157986Sdwmalone when a file is modified. Sendmail compensates by resetting the file mode 250171253Srwatson back to it's original settings. Unfortunately, there's still a 251157986Sdwmalone permission failure race as sendmail checks the permissions before locking 252157986Sdwmalone the file. This is unavoidable as sendmail must verify the file is safe 253157986Sdwmalone to open before opening it. A file can not be locked until it is open. 254157986Sdwmalone 255157986Sdwmalone* MAIL_HUB always takes precedence over LOCAL_RELAY 256157986Sdwmalone 257157986Sdwmalone Despite the information in the documentation, MAIL_HUB ($H) will always 258171253Srwatson be used if set instead of LOCAL_RELAY ($R). This will be fixed in a 259157986Sdwmalone future version. 260157986Sdwmalone 261101099Srwatson$Revision: 8.60 $, Last updated $Date: 2007/12/04 01:16:50 $ 262101099Srwatson