KNOWNBUGS revision 64562
1 2 3 K N O W N B U G S I N S E N D M A I L 4 (for 8.9.3) 5 6 7The following are bugs or deficiencies in sendmail that I am aware of 8but which have not been fixed in the current release. You probably 9want to get the most up to date version of this from ftp.sendmail.org 10in /pub/sendmail/KNOWNBUGS. For descriptions of bugs that have been 11fixed, see the file RELEASE_NOTES (in the root directory of the sendmail 12distribution). 13 14This list is not guaranteed to be complete. 15 16 17* Null bytes are not handled properly in headers. 18 19 Sendmail should handle full binary data. As it stands, it handles 20 all values in the body, but only 0x01-0x80 and 0xA0-0xFF in 21 the header. Notably missing is 0x00, which would require a major 22 restructuring of the code -- for example, almost no C library support 23 could be used to handle strings. 24 25* Duplicate error messages. 26 27 Sometimes identical, duplicate error messages can be generated. As 28 near as I can tell, this is rare and relatively innocuous. 29 30* $c (hop count) macro improperly set. 31 32 The $c macro is supposed to contain the current hop count, for use 33 when calling a mailer. This macro is initialized too early, and 34 is always zero (or the value of the -c command line flag, if any). 35 This macro will probably be removed entirely in a future release; 36 I don't believe there are any mailers left that require it. 37 38* \231 considered harmful. 39 40 Header addresses that have the \231 character (and possibly others 41 in the range \201 - \237) behave in odd and usually unexpected ways. 42 43* accept() problem on SVR4. 44 45 Apparently, the sendmail daemon loop (doing accept()s on the network) 46 can get into a weird state on SVR4; it starts logging ``SYSERR: 47 getrequests: accept: Protocol Error''. The workaround is to kill 48 and restart the sendmail daemon. We don't have an SVR4 system at 49 Berkeley that carries more than token mail load, so I can't validate 50 this. It is likely to be a glitch in the sockets emulation, since 51 "Protocol Error" is not possible error code with Berkeley TCP/IP. 52 53 I've also had someone report the message ``sendmail: accept: 54 SIOCGPGRP failed errno 22'' on an SVR4 system. This message is 55 not in the sendmail source code, so I assume it is also a bug 56 in the sockets emulation. (Errno 22 is EINVAL "Invalid Argument" 57 on all the systems I have available, including Solaris 2.x.) 58 Apparently, this problem is due to linking -lc before -lsocket; 59 if you are having this problem, check your Makefile. 60 61* accept() problem on Linux. 62 63 The accept() in sendmail daemon loop can return ETIMEDOUT. An 64 error is reported to syslog: 65 66 Jun 9 17:14:12 hostname sendmail[207]: NOQUEUE: SYSERR(root): 67 getrequests: accept: Connection timed out 68 69 "Connection timed out" is not documented as a valid return from 70 accept(2) and this was believed to be a bug in the Linux kernel. 71 Later information from the Linux kernel group states that Linux 72 2.0 kernels follow RFC1122 while sendmail follows the original BSD 73 (now POSIX 1003.1g draft) specification. The 2.1.X and later kernels 74 will follow the POSIX draft. 75 76* Excessive mailing list nesting can run out of file descriptors. 77 78 If you have a mailing list that includes lots of other mailing 79 lists, each of which has a separate owner, you can run out of 80 file descriptors. Each mailing list with a separate owner uses 81 one open file descriptor (prior to 8.6.6 it was three open 82 file descriptors per list). This is particularly egregious if 83 you have your connection cache set to be large. 84 85* Connection caching breaks if you pass the port number as an argument. 86 87 If you have a definition such as: 88 89 Mport, P=[IPC], F=kmDFMuX, S=11/31, R=21, 90 M=2100000, T=DNS/RFC822/SMTP, 91 A=IPC [127.0.0.1] $h 92 93 (i.e., where $h is the port number instead of the host name) the 94 connection caching code will break because it won't notice that 95 two messages addressed to different ports should use different 96 connections. 97 98* ESMTP SIZE underestimates the size of a message 99 100 Sendmail makes no allowance for headers that it adds, nor does it 101 account for the SMTP on-the-wire \r\n expansion. It probably doesn't 102 allow for 8->7 bit MIME conversions either. 103 104* Paths to programs being executed and the mode of program files are 105 not checked. Essentially, the RunProgramInUnsafeDirPath and 106 RunWritableProgram bits in the DontBlameSendmail option are always 107 set. This is not a problem if your system is well managed (that is, 108 if binaries and system directories are mode 755 instead of something 109 foolish like 777). 110 111* 8-bit data in GECOS field 112 113 If the GECOS (personal name) information in the passwd file contains 114 8-bit characters, those characters can be included in the message 115 header, which can cause problems when sending SMTP to hosts that 116 only accept 7-bit characters. 117 118* 8->7 bit MIME conversion 119 120 When sendmail is doing 8->7 bit MIME conversions, and the message 121 contains certain MIME body types that cannot be converted to 7-bit, 122 sendmail will strip the message to 7-bit. 123 124* 7->8 bit MIME conversion 125 126 If a message that is encoded as 7-bit MIME is converted to 8-bit and 127 that message when decoded is illegal (e.g., because of long lines or 128 illegal characters), sendmail can produce an illegal message. 129 130* MIME encoded full name phrases in the From: header 131 132 If a full name phrase includes characters from MustQuoteChars, sendmail 133 will quote the entire full name phrase. If MustQuoteChars includes 134 characters which are not special characters according to STD 11 (RFC 135 822), this quotation can interfere with MIME encoded full name phrases. 136 By default, sendmail includes the single quote character (') in 137 MustQuoteChars even though it is not listed as a special character in 138 STD 11. 139 140* bestmx map with -z flag truncates the list of MX hosts 141 142 A bestmx map configured with the -z flag will truncate the list 143 of MX hosts. This prevents creation of strings which are too 144 long for ruleset parsing. This can have an adverse effect on the 145 relay_based_on_MX feature. 146 147* Saving to ~sender/dead.letter fails if su'ed to root 148 149 If ErrorMode is set to print and an error in sending mail occurs, 150 the normal action is to print a message to the screen and append 151 the message to a dead.letter file in the sender's home directory. 152 In the case where the sender is using su to act as root, the file 153 safety checks prevent sendmail from saving the dead.letter file 154 because the sender's uid and the current real uid do not match. 155 156* Berkeley DB 2.X race condition with fcntl() locking 157 158 There is a race condition for Berkeley DB 2.X databases on 159 operating systems which use fcntl() style locking, such as 160 Solaris. Sendmail locks the map before calling db_open() to 161 prevent others from modifying the map while it is being opened. 162 Unfortunately, Berkeley DB opens the map, closes it, and then 163 reopens it. fcntl() locking drops the lock when any file 164 descriptor pointing to the file is closed, even if it is a 165 different file descriptor than the one used to initially lock 166 the file. As a result there is a possibility that entries in a 167 map might not be found during a map rebuild. As a workaround, 168 you can use makemap to build a map with a new name and then 169 "mv" the new db file to replace the old one. 170 171 Sleepycat Software has added code to avoid this race condition to 172 Berkeley DB versions after 2.7.5. 173 174* File open timeouts not available on hard mounted NFS file systems 175 176 Since SIGALRM does not interrupt an RPC call for hard mounted 177 NFS file systems, it is impossible to implement a timeout on a file 178 open operation. Therefore, while the NFS server is not responding, 179 attempts to open a file on that server will hang. Systems with 180 local mail delivery and NFS hard mounted home directories should be 181 avoided, as attempts to open the forward files could hang. 182 183* Race condition for delivery to setuid files 184 185 Sendmail will deliver to a fail if the file is owned by the DefaultUser 186 or has the setuid bit set. Unfortunately, some systems clear that bit 187 when a file is modified. Sendmail compensates by resetting the file mode 188 back to it's original settings. Unfortunately, there's still a 189 permission failure race as sendmail checks the permissions before locking 190 the file. This is unavoidable as sendmail must verify the file is safe 191 to open before opening it. A file can not be locked until it is open. 192 193* Potential denial of service attack with AutoRebuildAliases 194 195 There is a potential for a denial of service attack if the 196 AutoRebuildAliases option is set as a user can kill the sendmail process 197 while it is rebuilding the aliases file leaving it in an inconsistent 198 state. This option and it's use is deprecated and will be removed from a 199 future version of sendmail. 200 201$Revision: 8.43 $, Last updated $Date: 1999/11/17 18:56:09 $ 202