Andre Noll [Wed, 26 Jul 2017 18:45:27 +0000 (20:45 +0200)]
udp sender: Send EOF package only once.
We already have the per target ->sent_fec_eof flag, but we only set it
after the EOF packet was sent *successfully*. It's pointless to send
the packet more than once, so this patch modifies udp_close_target() to
set the flag regardless of whether the write(2) call succeeded.
Andre Noll [Mon, 24 Jul 2017 22:18:50 +0000 (00:18 +0200)]
server: Fix segfault in com_sender().
Commit 1709cc8f (server: Convert non-afs commands to lopsub) from
one year ago dropped the terminating NULL pointer from the array
of sender commands, presumably because the array size is declared
as the NUM_SENDER_CMDS enum constant, and this constant can be used
to iterate over all sender subcommands.
However, the loop in check_sender_args() of command.c does not
terminate the loop after NUM_SENDER_CMDS elements but only when it
encounters a NULL pointer. Hence, without the terminating NULL, the
code reads beyond the end of the array. The resulting invalid memory
access causes the command handler process to segfault.
Fix this by changing the termination condition of the loop to check
the loop variable against NUM_SENDER_CMDS.
Andre Noll [Mon, 24 Jul 2017 15:26:44 +0000 (17:26 +0200)]
vss: Avoid use after free in vss_send().
In vss_send() we increment the current slice number for the fec
client after fc->send_fec() has sent a slice. This results in a use
after free in case of a write error because ->send_fec() frees the
fec client structure on write errors. Valgrind complains about this
with the splat below.
To avoid this, the fc pointer must not be dereferenced after
->send_fec() has been called. This patch increases the current slice
number *before* the call to ->send_fec(). This works because the fec
clients do not care about this number.
The bug was introduced eight years ago in commit 625c5cd (Add forward
error correction code to the udp sender/receiver).
==8615== Invalid read of size 1
==8615== at 0x805022B: vss_send (vss.c:1051)
==8615== by 0x805022B: vss_post_select (vss.c:1168)
==8615== by 0x8061DC7: call_post_select (sched.c:84)
==8615== by 0x8061DC7: sched_post_select (sched.c:110)
==8615== by 0x8061DC7: schedule (sched.c:163)
==8615== by 0x804CBFD: main (server.c:607)
==8615== Address 0x4670168 is 80 bytes inside a block of size 116 free'd
==8615== at 0x402D221: free (vg_replace_malloc.c:530)
==8615== by 0x8062D7C: udp_delete_target (udp_send.c:80)
==8615== by 0x80630DC: udp_send_fec (udp_send.c:305)
==8615== by 0x805022A: vss_send (vss.c:1049)
==8615== by 0x805022A: vss_post_select (vss.c:1168)
==8615== by 0x8061DC7: call_post_select (sched.c:84)
==8615== by 0x8061DC7: sched_post_select (sched.c:110)
==8615== by 0x8061DC7: schedule (sched.c:163)
==8615== by 0x804CBFD: main (server.c:607)
==8615== Block was alloc'd at
==8615== at 0x402C1F0: malloc (vg_replace_malloc.c:299)
==8615== by 0x8052D7E: para_malloc (string.c:67)
==8615== by 0x8052FBD: para_calloc (string.c:90)
==8615== by 0x804F48F: vss_add_fec_client (vss.c:686)
==8615== by 0x8063433: udp_com_add (udp_send.c:342)
==8615== by 0x8063688: udp_init_target_list (udp_send.c:395)
==8615== by 0x806371D: udp_send_init (udp_send.c:442)
==8615== by 0x805062A: init_vss_task (vss.c:1195)
==8615== by 0x804CA57: server_init (server.c:537)
==8615== by 0x804CA57: main (server.c:605)
Andre Noll [Fri, 21 Jul 2017 17:45:31 +0000 (19:45 +0200)]
Merge branch 'refs/heads/t/doxygen'
This series contains an update of Doxyfile to a newer doxygen version
and an overhaul of the source code documentation. Several stale
references have been fixed. Other parts of the documentation have
been improved.
Cooking for almost a month.
* refs/heads/t/doxygen:
aft.c: Trivial spelling/whitespace fixes.
afh: Expand documentation of init function declarations.
filter: Remove duplicate documentation of filter_get().
doxygen: Improve documentation of struct receiver.
doxygen: Add \ref to references.
Improve documentation of mm.c and mm.h.
doxygen: Don't refer to Black Hats Manual.
doxygen: Don't refer to libosl functions.
doxygen: Trivial cleanups.
doxygen: Remove some stale doxygen references.
recv: Explain user data mechanism.
Update to doxygen-1.8.11.
The merge resulted in a few conflicts which were easy to resolve.
Andre Noll [Tue, 11 Jul 2017 17:33:31 +0000 (19:33 +0200)]
aft.c: Improve documentation of get_afhi_of_row().
It is not obvious that the function initializes the afh_info structure
so that its members point to memory-mapped data. This comment adds
some clarifying comments.
Andre Noll [Thu, 6 Jul 2017 18:46:29 +0000 (20:46 +0200)]
mood.c: Rename compute_mood_score().
The main purpose of this function is to determine whether the
audio file is admissible with respect to the given mood. So
row_is_admissible() is more to the point.
Andre Noll [Fri, 14 Jul 2017 13:39:28 +0000 (15:39 +0200)]
manual: libosl is optional.
Even without libosl, the build succeeds (only para_server won't be
built). Hence libosl should be listed as optional software. This
commit moves the item out of the list of mandatory packages.
Andre Noll [Sat, 8 Jul 2017 16:36:03 +0000 (18:36 +0200)]
build: Fix check for lopsub.
If the lopsub library is not installed, HAVE_LOPSUB should be set to
"no". The current code in configure.ac gets this wrong, causing the
configure script to succeed even if the library is not installed.
Andre Noll [Fri, 14 Jul 2017 12:50:43 +0000 (14:50 +0200)]
Merge branch 'refs/heads/t/manual'
A couple of patches which merge the contents of the overview.pdf file
into the user manual. After this, the documentation no longer depends
on dia.
Cooking for six weeks.
* refs/heads/t/manual:
manual: Expand description of para_play.
manual: Omit level 3 headers from table of contents.
manual: Combine overview.pdf and the user manual.
manual: Remove text about permissions of /var/paraslash.
manual: Correct format of para_client help output.
manual: Add lopsub installation instructions to quick start.
Andre Noll [Thu, 13 Jul 2017 15:21:13 +0000 (17:21 +0200)]
Merge branch 'maint'
A single important fix for the error subsystem. The merge conflicted
because master removed the E_AAC_AFH_INIT error code while maint
still has it. The conflict was trivial to resolve.
* refs/heads/maint:
Let error codes start out at index 1.
Andre Noll [Sat, 8 Jul 2017 20:55:11 +0000 (22:55 +0200)]
Let error codes start out at index 1.
When the error code facility was simplified one year ago in commit a775408 (Simplify the error subsystem, get rid of error2.[ch]) we
introduced a simple enum for all error codes, but missed the fact
that this approach assigns the first error code the value zero.
Hence, when the AAC decoder returns this error code, it will not be
recognized as an error because we generally check for negative return
values only.
This patch introduces an unused dummy error code so that the real
error codes start at index 1.
Andre Noll [Tue, 13 Jun 2017 19:48:18 +0000 (21:48 +0200)]
filter: Remove duplicate documentation of filter_get().
This public function is documented both in filter.h and in
filter_common.c. Get rid of the comment in the header file as this
is incomplete and causes a doxygen warning.
Andre Noll [Tue, 30 May 2017 19:11:09 +0000 (21:11 +0200)]
doxygen: Improve documentation of struct receiver.
The documentation still referred to the former gengetopt related
->conf pointer and also had some language and punctuation issues. The
new text should fix this, and is more to the point.
Andre Noll [Tue, 16 May 2017 21:09:33 +0000 (23:09 +0200)]
doxygen: Remove some stale doxygen references.
When we switched to lopsub, the callback request functions became
unused and were removed. At the same time, all command handlers were
made static, so they are no longer part of the doxygen documentation
and we shouldn't refer to them any more.
Also struct slot_info and the server_uptime variable have been made
private to their respective files audiod and time.c. Tthe latter was
in fact replaced by a timeval struct long ago.
The http and udp receivers don't have an init function any more, and
get_chunk_table_of_row() was removed ten years ago in commit 54a480ae.
Most references of the source code documentation will be annotated
with \ref in a subsequent commit to avoid this kind of documentation
bug in the future.
Andre Noll [Tue, 16 May 2017 21:16:44 +0000 (23:16 +0200)]
recv: Explain user data mechanism.
This is not obvious, so it deserves a comment in recv_init(). The
patch also adds a reference to each of the four instances of struct
receiver, reducing the number of doxygen warnings when EXTRACT_ALL
is set to NO in Doxyfile.
Andre Noll [Sun, 11 Jun 2017 10:13:22 +0000 (12:13 +0200)]
net.c: Silence clang warning.
clang-3.8.0 on FreeBSD complains about a recent change to net.c:
net.c:606:33: warning: missing field 'ss_family' initializer [-Wmissing-field-initializers]
struct sockaddr_storage ss = {0};
^
This line was introduced in commit 63128eea (net: Always initialize
struct sockaddr_storage) to avoid a warning from the static analyzer
of clang. Setting .ss_family initializes all struct members, which
makes both the analyzer and the compiler happy again.
Andre Noll [Tue, 13 Jun 2017 20:33:26 +0000 (22:33 +0200)]
filter: Handle bad filter numbers gracefully.
This modifies filter_get() to return NULL if an invalid filter number
is given, rather than aborting the process. This way applications
can loop over all filters without having to include filter_cmd.lsg.h.
filter_name(), which also receives a filter number, has no such
sanity check and could possibly access uninitialized memory if an
invalid argument was passed. This commit adds the check and makes the
function return NULL in this case, just like filter_get(). Moreover,
the function lacked documentation, so let's add it now.
Andre Noll [Tue, 13 Jun 2017 20:09:01 +0000 (22:09 +0200)]
audiod: Fix off-by-one bug in init_default_filters().
Filter numbers are one-based because zero is the number of the
(non-existing) supercommand of the filter_cmd suite. The loop in
init_default_filters() gets this wrong.
This bites only if no filter is given for at least one audio format.
Fortunately, the bug was easy to find because it triggers an assertion
in filter_get() which checks that the given number is at least one.
Andre Noll [Tue, 16 May 2017 20:21:12 +0000 (22:21 +0200)]
recv_common.c: Fix doxygen warning in recv_init().
In the doxygen comment of this function, one argument name was prefixed
with a backslash by mistake. Old doxygen versions did not complain
about this, but doxygen-1.8.11 does:
warning: unexpected token in comment block while parsing the argument of command param
Andre Noll [Mon, 1 May 2017 16:28:52 +0000 (18:28 +0200)]
manual: Expand description of para_play.
The text did not contain any characteristic features. Let's at least
mention that all audio formats are supported and that the command
line interface was designed to not suck.
Andre Noll [Sat, 29 Apr 2017 19:51:49 +0000 (21:51 +0200)]
manual: Omit level 3 headers from table of contents.
The table of contents was rather large because it contained too much
detail. This commit converts all level 3 headings to inline html which
has no visible effect except that the headings no longer appear in
the table of contents.
Andre Noll [Sat, 29 Apr 2017 02:33:17 +0000 (04:33 +0200)]
manual: Combine overview.pdf and the user manual.
The figures of the overview.pdf file look rather odd although they
used to look better when they were created four years ago. Apparently,
the dia command line tool which creates the pdf from the .dia source
file has changed and now messes up the lines which connect the parts
of the figure.
This commit re-implements four out of the five figures of the overview
file as ascii art in manual.md, removes overview.dia, and modifies
the documentation web page to no longer refer to the dia version.
The text of the overview file has been merged with the manual and
got improved a bit.
Andre Noll [Tue, 16 May 2017 20:08:19 +0000 (22:08 +0200)]
doxygen: Avoid segfault.
After an upgrade to Ubuntu-16.04, the new doxygen version 1.8.11
started to segfault. It turned out that this can be circumvented
by disabling USE_HTAGS so that doxygen will use its built-in source
browser instead of the htags tool of GNU global.
However, the html which is generated by the built-in source browser
the source code is unreadable because leading tab characters are not
shown, so indentation is messed up.
This patch deactivates the source browser for now, getting rid of
all source code in the generated output.
Andre Noll [Sat, 6 May 2017 15:40:38 +0000 (17:40 +0200)]
client: Silence a gcc-7.1,0 warning.
This version of gcc complains because a case statement in
client_pre_select() may fall through. The fallthrough is intentional,
however, and this is annotated with a comment. gcc-7 scans the comments
in a switch statement and suppresses the warning if it recognizes
a comment which says that fallthrough is intended. But for this to
happen, we must spell it correctly, and without whitespace.
After this commit, gcc-7 compiles the tree with no warnings.
Andre Noll [Sat, 29 Apr 2017 14:54:03 +0000 (16:54 +0200)]
Remove unused error code MPI_PRINT.
This had been removed already in commit eab41e87 (gcrypt: Remove
open-coded OAEP padding) but it was brought back by mistake in commit 9264932e (Merge branch 'refs/heads/t/dynamic_chunks').
Andre Noll [Thu, 27 Apr 2017 18:39:32 +0000 (20:39 +0200)]
Merge branch 'refs/heads/t/lopsub'
The bulk of the changes in this release is the conversion of all
command line parsers from gengetopt to lopsub.
The series also contains a few cleanups that have become possible
due to the switch from gengetopt to lopsub.
The patches towards the end of the series rename para_fade to
para_mixer.
Naturally, the merge conflicted rather heavily against the other
topic branches that have been merged since the lopsub branch was
started. Conflicting files:
The resolutions for these conflicts were recorded with git rerere
and have been tested for quite some time.
Cooking for three weeks.
* refs/heads/t/lopsub: (74 commits)
audioc: Avoid double free in audioc_i9e_line_handler().
audiod: Avoid uninitialized memory access.
Simplify mixer setup.
mixer: Implement non-linear time scale for fading.
mixer: Allow arbitrary relative time for sleep subcommand.
Convert para_fade to subcommands, rename it to para_mixer.
build: Create .dep files only during compilation.
build: Simplify definition of $m4_lls_deps.
build: Rename command list variables.
build: Combine $(CFLAGS) and $(STRICT_CFLAGS).
build: Let .d files depend only on .c.
build: Don't create phony targets for dependencies.
build: Remove duplicate dependency.
build: Remove cmdline_dir and friends.
build: Remove some unused variables from Makefile.real.
build: Remove m4/gengetopt.
Remove gengetopt and help2man checks from configure.ac.
Remove man_util.bash.
Remove ggo.c and ggo.h.
manual: Do not mention gengetopt and help2man any more.
...
Andre Noll [Fri, 21 Apr 2017 18:30:12 +0000 (20:30 +0200)]
vss: Add an assertion for vsst->header_buf.
if vsst->header_len > 0 but vsst->header_buf is NULL, something weird
is going on and non-obvious bad things will happen later. The assertion
makes sure we notice early that there is a problem. It also silences a
scan-build warning.
Andre Noll [Tue, 25 Apr 2017 16:04:57 +0000 (18:04 +0200)]
Merge branch 'maint'
A couple of fixes found by the clang static analyzer.
* maint:
aft: Check return value of load_afsi().
audiod: Avoid reading garbage in get_time_string().
net: Always initialize struct sockaddr_storage.
wma_afh: Fix two bugs in convert_utf8_to_utf16().
Andre Noll [Mon, 24 Apr 2017 18:05:18 +0000 (20:05 +0200)]
Merge branch 'refs/heads/t/xz'
A single patch which modifies the build system to create xz-compressed
tarballs. The patch conflicted against the changes introduced by the
recently merged makefile_conventions topic branch.
Cooking for almost a month.
* refs/heads/t/xz:
Switch to xz to compress tarballs.
Andre Noll [Sun, 23 Apr 2017 19:12:22 +0000 (21:12 +0200)]
Merge branch 'refs/heads/t/dynamic_chunks'
A rather long and intrusive series that finally improves the aac
decoder and audio format handler.
The merge had a minor conflict in error.h, which was easily resolved
by removing both affected error codes.
Cooking for a month.
* refs/heads/t/dynamic_chunks: (24 commits)
afh_recv: Check return value of afh_get_chunk().
aacdec: Decode only one frame per iteration.
aacdec: Combine aac_open() and aacdec_open().
aacdec: Make "initialized" a boolean.
aacdec: Make frame_info local to ->post_select().
aacdec: Remove pointless assignment.
aacdec: Remove pointless check and improve comment.
aacdec: Rename buffer variables.
aacdec: Remove pointless variable p.
aacdec: Remove check which is always true.
aacdec: Kill pointless label.
aacdec: Remove superfluous assignment.
aacdec: Improve and silence error message.
aacdec: Prefer NeAACDecInit() over NeAACDecInit2().
aacdec: Don't eat full buffer on errors.
Combine aacdec and aac_common.
Convert the aac audio format handler to libmp4ff.
aac_afh: Don't create chunk tables any more.
afh: Dynamic chunks.
server: Store max chunk size in database.
...
Andre Noll [Fri, 21 Apr 2017 18:57:38 +0000 (20:57 +0200)]
afh_recv: Check return value of afh_get_chunk().
Starting with commit b6b571e6c (afh: Dynamic chunks) the function
returns an int value and may fail. Although we store the return value
in a variable, the value of the variable is never read, causing the
clang static analyzer to complain (rightfully).
This patch modifies afh_recv_post_select() to fail if afh_get_chunk()
returns negative.
Andre Noll [Fri, 21 Apr 2017 20:54:36 +0000 (22:54 +0200)]
aft: Check return value of load_afsi().
This function should never fail, but it does return an error code and
all callers except copy_selector_info() check the return value. So
let's add the check there as well.
Andre Noll [Fri, 21 Apr 2017 18:43:06 +0000 (20:43 +0200)]
audiod: Avoid uninitialized memory access.
In handle_connect(), the local pointer variable errctx is not
initialized. If an error occurs early in the function, for example
because recv_cred_buffer() fails, we jump to the "out" label and try
to print the string errctx points to. This will cause a segfault or
worse, so make sure the variable is always initialized.
This bug was found by the static analyzer of clang.
Andre Noll [Wed, 19 Apr 2017 17:13:20 +0000 (19:13 +0200)]
Merge branch 'refs/heads/t/makefile_conventions'
A bunch of changes which strive to make the paraslash build system
more standard.
Cooking for three months.
* refs/heads/t/makefile_conventions:
build: Add target "check" as a synonym for "test".
build: Rename target "tarball" to "dist".
build: Do not strip installed executables by default.
build: Support $(DESTDIR).
build: Rename clean targets.
Andre Noll [Sun, 16 Apr 2017 18:15:35 +0000 (20:15 +0200)]
Merge branch 'refs/heads/t/rm_rc4'
This patch removes support for RC4, making the AES-based stream
cipher mandadory. The aes_ctr128 server feature is made a no-op,
breaking support with very old clients (<= 0.5.1).
Andre Noll [Sun, 9 Apr 2017 11:47:15 +0000 (13:47 +0200)]
Merge branch 'refs/heads/t/rm_asn'
A few cleanups, and one patch which removes an obsolete feature,
getting rid of an open-coded ASN parser.
Cooking for three months.
* refs/heads/t/rm_asn:
Remove unused E_PUBLIC_KEY.
crypt.c: Combine load_key() and get_private_key().
crypto: Remove support for ASN public keys.
crypto: Simplify asymetric key handling.
crypto: Rename check_key_file() -> check_private_key_file().
gcrypt.c: Always initialize result pointer in get_private_key().
Andre Noll [Fri, 30 Dec 2016 23:06:10 +0000 (00:06 +0100)]
Simplify mixer setup.
The ->init method of the two mixer implementations initialize the
members of the mixer structure. This structure contains only function
pointers. Their values, once set, never change.
Hence we may initialize the structure at compile time and make it
constant.
Andre Noll [Fri, 30 Dec 2016 22:25:11 +0000 (23:25 +0100)]
mixer: Implement non-linear time scale for fading.
For some sound hardware, changing the volume channel has little
effect for small values but even an increment by one is audible for
high values. For such hardware it is desirable to scale the volume
in a non-linear way.
This patch implements this feature. The new --fade-exponent receives
an argument to control the level of non-linearity. The default value
of zero results in old behavior: linear scaling.
For the implementation we need to compute (the inverse of) t ->
t^alpha, where alpha is between zero and one. This requires the use
of exp(3) and log(3), so we need to include <math.h> and link with -lm.
Andre Noll [Sun, 3 Jul 2016 14:47:13 +0000 (16:47 +0200)]
Switch to xz to compress tarballs.
xz compresses better than bzip2, and decompression needs less time. It
is also widely used by many other software projects, so there should
be no disadvantage.
We keep removing the .bz2 tarball on make clean, and also keep the
paraslash-*.tar.bz2 entry in .gitignore for the time being.
Andre Noll [Tue, 27 Dec 2016 23:53:29 +0000 (00:53 +0100)]
mixer: Allow arbitrary relative time for sleep subcommand.
If --wake-hour is not given, the wake time defaults to now + nine
hours. This is arbitrary, and the constant 9 was hardcoded with no
way to specify it at the command line or in the config file.
This commit replaces --wake-hour and --wake-min by --wake-time whose
argument must be given in HH:MM syntax. An optional leading '+'
character allows to specify a relative offset.
Andre Noll [Tue, 27 Dec 2016 21:19:03 +0000 (22:19 +0100)]
Convert para_fade to subcommands, rename it to para_mixer.
para_fade operates in one of several modes, but the command line
options are shared. So most of the options are ignored for any
particular mode. With lopsub we can do better and define one
subcommand per mode so that options which are only relevant to one
subcommand can be made local to that subcommand.
With subcommands the syntax of the command changes, which breaks
existing scripts. So it seems to be prudent to rename the executable
from para_fade to para_mixer. The old name was a bit misleading anyway
since the command can do much more than just fade the volume.
Andre Noll [Sat, 21 Jan 2017 15:18:44 +0000 (16:18 +0100)]
build: Create .dep files only during compilation.
It is unnecessary to run a dedicated command to create the dependency
files because:
* If the object file does not exist, we have to run the compiler
anyway, which creates the dependency on the fly.
* If the object file exists, the dependency file also exists, so
the build system notices if the object file needs to be rebuilt,
for example because an included header file was modified.
The only thing we must guarantee is that all included header files
exist. We do this by introducing an order-only dependency: all object
files depend on all .lsg.h files but we impose this specific ordering
without forcing the targets to be updated.
This speeds up the build a lot and also simplifies Makefile.real.
Andre Noll [Sun, 17 Jul 2016 18:54:00 +0000 (20:54 +0200)]
build: Rename command list variables.
The old names, xxx_command_lists and man_util_command_lists are not
accurate any more, and unnecessarily clumsy to boot.
This patch renames the variables to xxx_commands and all_commands and
sorts the three affected sections of the Makefile alphabetically to
reduce the risk of future merge conflicts.
Andre Noll [Sun, 17 Jul 2016 16:27:13 +0000 (18:27 +0200)]
build: Combine $(CFLAGS) and $(STRICT_CFLAGS).
We used to compile the parsers generated by gengetopt with less strict
flags to avoid warnings in generated files. With lopsub this is no
longer necessary, so one set of flags suffices.
Moreover, we currently initialize $(STRICT_CFLAGS) to $(CFLAGS) (from
the environment), then append our own flags. This is unfortunate
because the user defined settings in $(CFLAGS) are overwritten.
This patch changes the Makefile to only set STRICT_CFLAGS. We now
pass $(STRICT_CFLAGS) $(CFLAGS) to the compiler, allowing the user
to override (parts of) $(STRICT_CFLAGS) by adding -Wno-foo to CFLAGS.
We use the opportunity to regroup the various options a little.
Andre Noll [Sat, 16 Jul 2016 20:24:12 +0000 (22:24 +0200)]
build: Let .d files depend only on .c.
Currently we let each .d file depend on the same set of files as the
.o file by having two targets specified as arguments to -MT options
in the cc command that creates the .d file: one for the .d file and
one for the .o file.
This approach runs into problems if header files are removed and the
now stale .d file depends on a file which no longer exists.
It is enough to let the .d file depend only on the corresponding
.c file. In case an included file changes or is removed, the .c
file changes as well, so we already know we must build the object
file and can update the .d file at the same time, without running
cc twice. The -MMD option which is added to the usual cc command
instructs the compiler to create both the .d and the .o file in one go.
When creating the .o and .d files simultaneously, we don't need a
sed command to prepend the directory containing the .lsg.h headers
as these files are prerequisites of the object file being created,
so they are known to exist when the cc command is executed.
Andre Noll [Sat, 16 Jul 2016 11:43:52 +0000 (13:43 +0200)]
build: Don't create phony targets for dependencies.
The -MP option is supposed to work around make errors after a header
file was removed. However, with the option enabled, the attempt to
include a non-existing header file causes the build to loop endlessly.
Andre Noll [Sun, 10 Jul 2016 16:33:48 +0000 (18:33 +0200)]
build: Remove cmdline_dir and friends.
This directory used to contain the source files of the command line
parsers generated by gengetopt. There are no such files any more,
so the the receipts which compile these source files can go away and
we do not need to create the directory any more.
The special CFLAGS for compiling those files can be removed as well
as the include directive for CPPFLAGS.
Finally, the make construct which filters out the .cmdline.d files
from the list of dependency files is no longer needed, and we may get
rid of the sed command which prepends $(cmdline_dir) to the includes
of the dependency files.
Andre Noll [Wed, 29 Jun 2016 20:21:35 +0000 (22:21 +0200)]
build: Remove some unused variables from Makefile.real.
All executables are converted, so $converted_executables is the same
as $executables, modulo executables which can not be built (and are
thus irrelevant). Similarly, $unconverted_executables is empty and
can be removed as well.
The $m4_deps variable was only used for the gengetopt files, so this
can also be removed, along with $ggo_dir.
$help2man_dir used to contain the path to the directory for the output
files of help2man. Since we now generate the man pages with lopsubgen,
$help2man_dir has become unused and can be removed.