3 K N O W N B U G S I N S E N D M A I L
6 The following are bugs or deficiencies in sendmail that we are aware of
7 but which have not been fixed in the current release. You probably
8 want to get the most up to date version of this from ftp.sendmail.org
9 in /pub/sendmail/KNOWNBUGS. For descriptions of bugs that have been
10 fixed, see the file RELEASE_NOTES (in the root directory of the sendmail
13 This list is not guaranteed to be complete.
15 * Delivery to programs that generate too much output may cause problems
17 If e-mail is delivered to a program which generates too much
18 output, then sendmail may issue an error:
20 timeout waiting for input from local during Draining Input
22 Make sure that the program does not generate output beyond a
23 status message (corresponding to the exit status). This may
24 require a wrapper around the actual program to redirect output
27 Such a problem has been reported for bulk_mailer.
29 * Null bytes are not handled properly in headers.
31 Sendmail should handle full binary data. As it stands, it handles
32 all values in the body, but not 0x00 in the header. Changing
33 this would require a major restructuring of the code -- for
34 example, almost no C library support could be used to handle
37 * Header checks are not called if header value is too long or empty.
39 If the value of a header is longer than 1250 (MAXNAME + MAXATOM - 6)
40 characters or it contains a single word longer than 256 (MAXNAME)
41 characters then no header check is done even if one is configured for
44 * Header lines which are too long will be split incorrectly.
46 Header lines which are longer than 2045 characters will be split
47 but some characters might be lost. Fix: obey RFC (2)822 and do not
48 send lines that are longer than 1000 characters.
50 * milter communication fails if a single header is larger than 64K.
52 If a single header is larger than 64KB (which is not possible in the
53 default configuration) then it cannot be transferred in one block to
54 libmilter and hence the communication fails. This can be avoided by
55 increasing the constant MILTER_CHUNK_SIZE in
56 include/libmilter/mfdef.h and recompiling sendmail, libmilter, and
57 all (statically linked) milters (or by using an undocumented compile
58 time option: _FFR_MAXDATASIZE; you have to read the source code in
59 order to use this properly).
61 * Sender addresses whose domain part cause a temporary A record lookup
62 failure but have a valid MX record will be temporarily rejected in
63 the default configuration. Solution: fix the DNS at the sender side.
64 If that's not easy to achieve, possible workarounds are:
65 - add an entry to the access map:
67 - (only for advanced users) replace
69 # Resolve map (to check if a host exists in check_mail)
70 Kresolve host -a<OKR> -T<TEMP>
74 # Resolve map (to check if a host exists in check_mail)
75 Kcanon host -a<OKR> -T<TEMP>
76 Kdnsmx dns -R MX -a<OKR> -T<TEMP>
77 Kresolve sequence dnsmx canon
80 * Duplicate error messages.
82 Sometimes identical, duplicate error messages can be generated. As
83 near as I can tell, this is rare and relatively innocuous.
85 * Misleading error messages.
87 If an illegal address is specified on the command line together
88 with at least one valid address and PostmasterCopy is set, the
89 DSN does not contain the illegal address, but only the valid
92 * \231 considered harmful.
94 Header addresses that have the \231 character (and possibly others
95 in the range \201 - \237) behave in odd and usually unexpected ways.
97 * accept() problem on SVR4.
99 Apparently, the sendmail daemon loop (doing accept()s on the network)
100 can get into a weird state on SVR4; it starts logging ``SYSERR:
101 getrequests: accept: Protocol Error''. The workaround is to kill
102 and restart the sendmail daemon. We don't have an SVR4 system at
103 Berkeley that carries more than token mail load, so I can't validate
104 this. It is likely to be a glitch in the sockets emulation, since
105 "Protocol Error" is not possible error code with Berkeley TCP/IP.
107 I've also had someone report the message ``sendmail: accept:
108 SIOCGPGRP failed errno 22'' on an SVR4 system. This message is
109 not in the sendmail source code, so I assume it is also a bug
110 in the sockets emulation. (Errno 22 is EINVAL "Invalid Argument"
111 on all the systems I have available, including Solaris 2.x.)
112 Apparently, this problem is due to linking -lc before -lsocket;
113 if you are having this problem, check your Makefile.
115 * accept() problem on Linux.
117 The accept() in sendmail daemon loop can return ETIMEDOUT. An
118 error is reported to syslog:
120 Jun 9 17:14:12 hostname sendmail[207]: NOQUEUE: SYSERR(root):
121 getrequests: accept: Connection timed out
123 "Connection timed out" is not documented as a valid return from
124 accept(2) and this was believed to be a bug in the Linux kernel.
125 Later information from the Linux kernel group states that Linux
126 2.0 kernels follow RFC1122 while sendmail follows the original BSD
127 (now POSIX 1003.1g draft) specification. The 2.1.X and later kernels
128 will follow the POSIX draft.
130 * Excessive mailing list nesting can run out of file descriptors.
132 If you have a mailing list that includes lots of other mailing
133 lists, each of which has a separate owner, you can run out of
134 file descriptors. Each mailing list with a separate owner uses
135 one open file descriptor (prior to 8.6.6 it was three open
136 file descriptors per list). This is particularly egregious if
137 you have your connection cache set to be large.
139 * Connection caching breaks if you pass the port number as an argument.
141 If you have a definition such as:
143 Mport, P=[IPC], F=kmDFMuX, S=11/31, R=21,
144 M=2100000, T=DNS/RFC822/SMTP,
147 (i.e., where $h is the port number instead of the host name) the
148 connection caching code will break because it won't notice that
149 two messages addressed to different ports should use different
152 * ESMTP SIZE underestimates the size of a message
154 Sendmail makes no allowance for headers that it adds, nor does it
155 account for the SMTP on-the-wire \r\n expansion. It probably doesn't
156 allow for 8->7 bit MIME conversions either.
158 * Client ignores SIZE parameter.
160 When sendmail acts as client and the server specifies a limit
161 for the mail size, sendmail will ignore this and try to send the
162 mail anyway. The server will usually reject the MAIL command
163 which specifies the size of the message and hence this problem
166 * Paths to programs being executed and the mode of program files are
167 not checked. Essentially, the RunProgramInUnsafeDirPath and
168 RunWritableProgram bits in the DontBlameSendmail option are always
169 set. This is not a problem if your system is well managed (that is,
170 if binaries and system directories are mode 755 instead of something
173 * 8-bit data in GECOS field
175 If the GECOS (personal name) information in the passwd file contains
176 8-bit characters, those characters can be included in the message
177 header, which can cause problems when sending SMTP to hosts that
178 only accept 7-bit characters.
180 * 8->7 bit MIME conversion
182 When sendmail is doing 8->7 bit MIME conversions, and the message
183 contains certain MIME body types that cannot be converted to 7-bit,
184 sendmail will pass the message as 8-bit.
186 * 7->8 bit MIME conversion
188 If a message that is encoded as 7-bit MIME is converted to 8-bit and
189 that message when decoded is illegal (e.g., because of long lines or
190 illegal characters), sendmail can produce an illegal message.
192 * MIME encoded full name phrases in the From: header
194 If a full name phrase includes characters from MustQuoteChars, sendmail
195 will quote the entire full name phrase. If MustQuoteChars includes
196 characters which are not special characters according to STD 11 (RFC
197 822), this quotation can interfere with MIME encoded full name phrases.
198 By default, sendmail includes the single quote character (') in
199 MustQuoteChars even though it is not listed as a special character in
202 * bestmx map with -z flag truncates the list of MX hosts
204 A bestmx map configured with the -z flag will truncate the list
205 of MX hosts. This prevents creation of strings which are too
206 long for ruleset parsing. This can have an adverse effect on the
207 relay_based_on_MX feature.
209 * Saving to ~sender/dead.letter fails if su'ed to root
211 If ErrorMode is set to print and an error in sending mail occurs,
212 the normal action is to print a message to the screen and append
213 the message to a dead.letter file in the sender's home directory.
214 In the case where the sender is using su to act as root, the file
215 safety checks prevent sendmail from saving the dead.letter file
216 because the sender's uid and the current real uid do not match.
218 * Berkeley DB 2.X race condition with fcntl() locking
220 There is a race condition for Berkeley DB 2.X databases on
221 operating systems which use fcntl() style locking, such as
222 Solaris. Sendmail locks the map before calling db_open() to
223 prevent others from modifying the map while it is being opened.
224 Unfortunately, Berkeley DB opens the map, closes it, and then
225 reopens it. fcntl() locking drops the lock when any file
226 descriptor pointing to the file is closed, even if it is a
227 different file descriptor than the one used to initially lock
228 the file. As a result there is a possibility that entries in a
229 map might not be found during a map rebuild. As a workaround,
230 you can use makemap to build a map with a new name and then
231 "mv" the new db file to replace the old one.
233 Sleepycat Software has added code to avoid this race condition to
234 Berkeley DB versions after 2.7.5.
236 * File open timeouts not available on hard mounted NFS file systems
238 Since SIGALRM does not interrupt an RPC call for hard mounted
239 NFS file systems, it is impossible to implement a timeout on a file
240 open operation. Therefore, while the NFS server is not responding,
241 attempts to open a file on that server will hang. Systems with
242 local mail delivery and NFS hard mounted home directories should be
243 avoided, as attempts to open the forward files could hang.
245 * Race condition for delivery to set-user-ID files
247 Sendmail will deliver to a fail if the file is owned by the DefaultUser
248 or has the set-user-ID bit set. Unfortunately, some systems clear that bit
249 when a file is modified. Sendmail compensates by resetting the file mode
250 back to it's original settings. Unfortunately, there's still a
251 permission failure race as sendmail checks the permissions before locking
252 the file. This is unavoidable as sendmail must verify the file is safe
253 to open before opening it. A file can not be locked until it is open.
255 * MAIL_HUB always takes precedence over LOCAL_RELAY
257 Despite the information in the documentation, MAIL_HUB ($H) will always
258 be used if set instead of LOCAL_RELAY ($R). This will be fixed in a
261 $Revision: 8.60 $, Last updated $Date: 2007/12/04 01:16:50 $