Message ID | 1251481751-27326-1-git-send-email-quintela@redhat.com |
---|---|
State | Superseded |
Headers | show |
On Fri, 28 Aug 2009, Juan Quintela wrote: > HAS_YMF262 is not defined anywhere, and use functinos that are not > in the tree. It doesn't hurt anyone, and should stay for those who want OPL3. [..snip..]
malc <av1474@comtv.ru> wrote: > On Fri, 28 Aug 2009, Juan Quintela wrote: > >> HAS_YMF262 is not defined anywhere, and use functinos that are not >> in the tree. > > It doesn't hurt anyone, and should stay for those who want OPL3. > > [..snip..] We don't have the functions that they protect from. Later, Juan.
On Fri, 28 Aug 2009, Juan Quintela wrote: > malc <av1474@comtv.ru> wrote: > > On Fri, 28 Aug 2009, Juan Quintela wrote: > > > >> HAS_YMF262 is not defined anywhere, and use functinos that are not > >> in the tree. > > > > It doesn't hurt anyone, and should stay for those who want OPL3. > > > > [..snip..] > > We don't have the functions that they protect from. What?
malc <av1474@comtv.ru> wrote: > On Fri, 28 Aug 2009, Juan Quintela wrote: > >> malc <av1474@comtv.ru> wrote: >> > On Fri, 28 Aug 2009, Juan Quintela wrote: >> > >> >> HAS_YMF262 is not defined anywhere, and use functinos that are not >> >> in the tree. >> > >> > It doesn't hurt anyone, and should stay for those who want OPL3. >> > >> > [..snip..] >> >> We don't have the functions that they protect from. > diff --git a/hw/adlib.c b/hw/adlib.c > index e1e4077..8b77ebd 100644 > --- a/hw/adlib.c > +++ b/hw/adlib.c > @@ -42,14 +42,8 @@ > #define ldebug(...) > #endif > > -#ifdef HAS_YMF262 > -#include "ymf262.h" $ find . -type f -name "ymf262.h" $ > -void YMF262UpdateOneQEMU (int which, INT16 *dst, int length); $ find . -type f -name "*.[ch]" | xargs grep YMF262UpdateOneQEMU ./hw/adlib.c:void YMF262UpdateOneQEMU (int which, INT16 *dst, int length); ./hw/adlib.c: YMF262UpdateOneQEMU (0, s->mixbuf + s->pos * 2, samples); $ > - YMF262TimerOver (0, n); $ find . -type f -name "*.[ch]" | xargs grep YMF262TimerOver ./hw/adlib.c: YMF262TimerOver (0, n); $ > - status = YMF262Write (0, a, val); $ find . -type f -name "*.[ch]" | xargs grep YMF262Write ./hw/adlib.c: status = YMF262Write (0, a, val); $ I.e. We don't have defined that functions anywhere in the source code. Are I missing something obvious? Later, Juan.
On Sat, 29 Aug 2009, Juan Quintela wrote: > malc <av1474@comtv.ru> wrote: > > On Fri, 28 Aug 2009, Juan Quintela wrote: > > > >> malc <av1474@comtv.ru> wrote: > >> > On Fri, 28 Aug 2009, Juan Quintela wrote: > >> > > >> >> HAS_YMF262 is not defined anywhere, and use functinos that are not > >> >> in the tree. > >> > > >> > It doesn't hurt anyone, and should stay for those who want OPL3. > >> > > >> > [..snip..] > >> > >> We don't have the functions that they protect from. > > [..snip..] > > I.e. We don't have defined that functions anywhere in the source code. > Are I missing something obvious? It doesn't exist inside the tree due to ymf262.c being under MAME license which is [L]GPL incompatible, i.e. the combination is non distributable, which does not, however, preclude the user from getting ymf262 elsehwere and building QEMU on his/her own.
malc <av1474@comtv.ru> wrote: > On Sat, 29 Aug 2009, Juan Quintela wrote: > >> malc <av1474@comtv.ru> wrote: >> > On Fri, 28 Aug 2009, Juan Quintela wrote: >> > >> >> malc <av1474@comtv.ru> wrote: >> >> > On Fri, 28 Aug 2009, Juan Quintela wrote: >> >> > >> >> >> HAS_YMF262 is not defined anywhere, and use functinos that are not >> >> >> in the tree. >> >> > >> >> > It doesn't hurt anyone, and should stay for those who want OPL3. >> >> > >> >> > [..snip..] >> >> >> >> We don't have the functions that they protect from. >> >> > [..snip..] >> >> I.e. We don't have defined that functions anywhere in the source code. >> Are I missing something obvious? > > It doesn't exist inside the tree due to ymf262.c being under MAME license > which is [L]GPL incompatible, i.e. the combination is non distributable, > which does not, however, preclude the user from getting ymf262 elsehwere > and building QEMU on his/her own. I am not a Lawyer, but that looks fishy at least. Later, Juan.
diff --git a/hw/adlib.c b/hw/adlib.c index e1e4077..8b77ebd 100644 --- a/hw/adlib.c +++ b/hw/adlib.c @@ -42,14 +42,8 @@ #define ldebug(...) #endif -#ifdef HAS_YMF262 -#include "ymf262.h" -void YMF262UpdateOneQEMU (int which, INT16 *dst, int length); -#define SHIFT 2 -#else #include "fmopl.h" #define SHIFT 1 -#endif #define IO_READ_PROTO(name) \ uint32_t name (void *opaque, uint32_t nport) @@ -75,20 +69,14 @@ typedef struct { SWVoiceOut *voice; int left, pos, samples; QEMUAudioTimeStamp ats; -#ifndef HAS_YMF262 FM_OPL *opl; -#endif } AdlibState; static AdlibState glob_adlib; static void adlib_stop_opl_timer (AdlibState *s, size_t n) { -#ifdef HAS_YMF262 - YMF262TimerOver (0, n); -#else OPLTimerOver (s->opl, n); -#endif s->ticking[n] = 0; } @@ -126,11 +114,7 @@ static IO_WRITE_PROTO (adlib_write) adlib_kill_timers (s); -#ifdef HAS_YMF262 - status = YMF262Write (0, a, val); -#else status = OPLWrite (s->opl, a, val); -#endif } static IO_READ_PROTO (adlib_read) @@ -141,11 +125,7 @@ static IO_READ_PROTO (adlib_read) adlib_kill_timers (s); -#ifdef HAS_YMF262 - data = YMF262Read (0, a); -#else data = OPLRead (s->opl, a); -#endif return data; } @@ -235,11 +215,7 @@ static void adlib_callback (void *opaque, int free) return; } -#ifdef HAS_YMF262 - YMF262UpdateOneQEMU (0, s->mixbuf + s->pos * 2, samples); -#else YM3812UpdateOne (s->opl, s->mixbuf + s->pos, samples); -#endif while (samples) { written = write_audio (s, samples); @@ -258,14 +234,10 @@ static void adlib_callback (void *opaque, int free) static void Adlib_fini (AdlibState *s) { -#ifdef HAS_YMF262 - YMF262Shutdown (); -#else if (s->opl) { OPLDestroy (s->opl); s->opl = NULL; } -#endif if (s->mixbuf) { qemu_free (s->mixbuf); @@ -281,16 +253,6 @@ int Adlib_init (qemu_irq *pic) AdlibState *s = &glob_adlib; struct audsettings as; -#ifdef HAS_YMF262 - if (YMF262Init (1, 14318180, conf.freq)) { - dolog ("YMF262Init %d failed\n", conf.freq); - return -1; - } - else { - YMF262SetTimerHandler (0, timer_handler, 0); - s->enabled = 1; - } -#else s->opl = OPLCreate (OPL_TYPE_YM3812, 3579545, conf.freq); if (!s->opl) { dolog ("OPLCreate %d failed\n", conf.freq); @@ -300,7 +262,6 @@ int Adlib_init (qemu_irq *pic) OPLSetTimerHandler (s->opl, timer_handler, 0); s->enabled = 1; } -#endif as.freq = conf.freq; as.nchannels = SHIFT; diff --git a/vl.c b/vl.c index a894285..8c53e25 100644 --- a/vl.c +++ b/vl.c @@ -4443,11 +4443,7 @@ struct soundhw soundhw[] = { #ifdef CONFIG_ADLIB { "adlib", -#ifdef HAS_YMF262 - "Yamaha YMF262 (OPL3)", -#else "Yamaha YM3812 (OPL2)", -#endif 0, 1, { .init_isa = Adlib_init }
HAS_YMF262 is not defined anywhere, and use functinos that are not in the tree. Signed-off-by: Juan Quintela <quintela@redhat.com> --- hw/adlib.c | 39 --------------------------------------- vl.c | 4 ---- 2 files changed, 0 insertions(+), 43 deletions(-)