MinGW fixes (development)
commitb26bdd3c19dce85a06839b34f36834cb5bb0330f
authorCary R <cygcary@yahoo.com>
Thu, 22 May 2008 23:31:09 +0000 (22 16:31 -0700)
committerStephen Williams <steve@icarus.com>
Fri, 23 May 2008 03:24:21 +0000 (22 20:24 -0700)
tree1312351d9580ba1bb9bb0c0f3f7e7256d66a220f
parent6a9c5d8f7e1e3b424c2bccc25419162631e4bf35
MinGW fixes (development)

The MinGW system() implementation appears to return the straight
return value instead of the waitpid() like result that more
normal systems return. Because of this just return the system()
result without processing for MinGW compilations.

Older version of the MinGW runtime (pre 3.14) just used the
underlying vsnprintf(). Which has some problems. The 3.14 version
has some nice improvements, but it has a sever bug when processing
"%*.*f", -1, -1, <some_real_value>.  Because of this we need to use
the underlying version without the enhancements for now.

snprintf prints %p differently than the other printf routines
so use _snprintf to get consistent results.

Only build the PDF files if both man and ps2pdf exist.

MinGW does not know about the z modifier for %d, %u, etc.

Add some missing Makefile check targets.
17 files changed:
Makefile.in
cadpli/Makefile.in
configure.in
driver-vpi/Makefile.in
driver/Makefile.in
driver/main.c
ivlpp/lexor.lex
tgt-null/Makefile.in
tgt-stub/Makefile.in
tgt-vvp/draw_vpi.c
tgt-vvp/modpath.c
tgt-vvp/vvp_scope.c
vpi/Makefile.in
vvp/Makefile.in
vvp/configure.in
vvp/udp.cc
vvp/vpi_mcd.cc