From: Paolo Bonzini <pbonzini@redhat.com>
To: qemu-devel@nongnu.org
Cc: peter.maydell@linaro.org
Subject: [Qemu-devel] [PATCH 1/3] i8257: rewrite DMA_schedule to avoid hooking into the CPU loop
Date: Fri, 14 Aug 2015 15:15:27 +0200 [thread overview]
Message-ID: <1439558129-466-2-git-send-email-pbonzini@redhat.com> (raw)
In-Reply-To: <1439558129-466-1-git-send-email-pbonzini@redhat.com>
The i8257 DMA controller uses an idle bottom half, which by default
does not cause the main loop to exit. Therefore, the DMA_schedule
function is there to ensure that the CPU relinquishes the iothread
mutex to the iothread.
However, this is not enough since the iothread will call
aio_compute_timeout() and go to sleep again. In the iothread
world, forcing execution of the idle bottom half is much simpler,
and only requires a call to qemu_notify_event(). Do it, removing
the need for the "cpu_request_exit" pseudo-irq. The next patch
will remove it.
Signed-off-by: Paolo Bonzini <pbonzini@redhat.com>
---
hw/block/fdc.c | 2 +-
hw/dma/i8257.c | 18 ++++++++++++------
hw/sparc/sun4m.c | 2 +-
hw/sparc64/sun4u.c | 2 +-
include/hw/isa/isa.h | 2 +-
5 files changed, 16 insertions(+), 10 deletions(-)
diff --git a/hw/block/fdc.c b/hw/block/fdc.c
index 5e1b67e..6686a72 100644
--- a/hw/block/fdc.c
+++ b/hw/block/fdc.c
@@ -1417,7 +1417,7 @@ static void fdctrl_start_transfer(FDCtrl *fdctrl, int direction)
* recall us...
*/
DMA_hold_DREQ(fdctrl->dma_chann);
- DMA_schedule(fdctrl->dma_chann);
+ DMA_schedule();
} else {
/* Start transfer */
fdctrl_transfer_handler(fdctrl, fdctrl->dma_chann, 0,
diff --git a/hw/dma/i8257.c b/hw/dma/i8257.c
index a414029..409ba7d 100644
--- a/hw/dma/i8257.c
+++ b/hw/dma/i8257.c
@@ -358,6 +358,7 @@ static void channel_run (int ncont, int ichan)
}
static QEMUBH *dma_bh;
+static bool dma_bh_scheduled;
static void DMA_run (void)
{
@@ -390,12 +391,15 @@ static void DMA_run (void)
running = 0;
out:
- if (rearm)
+ if (rearm) {
qemu_bh_schedule_idle(dma_bh);
+ dma_bh_scheduled = true;
+ }
}
static void DMA_run_bh(void *unused)
{
+ dma_bh_scheduled = false;
DMA_run();
}
@@ -458,12 +462,14 @@ int DMA_write_memory (int nchan, void *buf, int pos, int len)
return len;
}
-/* request the emulator to transfer a new DMA memory block ASAP */
-void DMA_schedule(int nchan)
+/* request the emulator to transfer a new DMA memory block ASAP (even
+ * if the idle bottom half would not have exited the iothread yet).
+ */
+void DMA_schedule(void)
{
- struct dma_cont *d = &dma_controllers[nchan > 3];
-
- qemu_irq_pulse(*d->cpu_request_exit);
+ if (dma_bh_scheduled) {
+ qemu_notify_event();
+ }
}
static void dma_reset(void *opaque)
diff --git a/hw/sparc/sun4m.c b/hw/sparc/sun4m.c
index 68ac4d8..ebaae9d 100644
--- a/hw/sparc/sun4m.c
+++ b/hw/sparc/sun4m.c
@@ -109,7 +109,7 @@ int DMA_write_memory (int nchan, void *buf, int pos, int size)
}
void DMA_hold_DREQ (int nchan) {}
void DMA_release_DREQ (int nchan) {}
-void DMA_schedule(int nchan) {}
+void DMA_schedule(void) {}
void DMA_init(int high_page_enable, qemu_irq *cpu_request_exit)
{
diff --git a/hw/sparc64/sun4u.c b/hw/sparc64/sun4u.c
index 30cfa0e..44eb4eb 100644
--- a/hw/sparc64/sun4u.c
+++ b/hw/sparc64/sun4u.c
@@ -112,7 +112,7 @@ int DMA_write_memory (int nchan, void *buf, int pos, int size)
}
void DMA_hold_DREQ (int nchan) {}
void DMA_release_DREQ (int nchan) {}
-void DMA_schedule(int nchan) {}
+void DMA_schedule(void) {}
void DMA_init(int high_page_enable, qemu_irq *cpu_request_exit)
{
diff --git a/include/hw/isa/isa.h b/include/hw/isa/isa.h
index f21ceaa..81b94ea 100644
--- a/include/hw/isa/isa.h
+++ b/include/hw/isa/isa.h
@@ -112,7 +112,7 @@ int DMA_read_memory (int nchan, void *buf, int pos, int size);
int DMA_write_memory (int nchan, void *buf, int pos, int size);
void DMA_hold_DREQ (int nchan);
void DMA_release_DREQ (int nchan);
-void DMA_schedule(int nchan);
+void DMA_schedule(void);
void DMA_init(int high_page_enable, qemu_irq *cpu_request_exit);
void DMA_register_channel (int nchan,
DMA_transfer_handler transfer_handler,
--
2.4.3
next prev parent reply other threads:[~2015-08-14 13:15 UTC|newest]
Thread overview: 6+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-08-14 13:15 [Qemu-devel] [PATCH 0/3] Signal-free qemu_cpu_kick for TCG Paolo Bonzini
2015-08-14 13:15 ` Paolo Bonzini [this message]
2015-08-14 13:15 ` [Qemu-devel] [PATCH 2/3] i8257: remove cpu_request_exit irq Paolo Bonzini
2015-08-14 13:15 ` [Qemu-devel] [PATCH 3/3] tcg: signal-free qemu_cpu_kick Paolo Bonzini
2015-08-17 18:31 ` Richard Henderson
2015-08-17 20:47 ` Paolo Bonzini
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=1439558129-466-2-git-send-email-pbonzini@redhat.com \
--to=pbonzini@redhat.com \
--cc=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).