* Re: [PATCH] change acquire/release_console_sem() to console_lock/unlock()
From: torbenh @ 2011-01-20 16:55 UTC (permalink / raw)
To: Greg KH
Cc: xen-devel, Kees Cook, linux-fbdev, Sascha Hauer, James Hogan,
David Airlie, Benjamin Herrenschmidt, Jesse Barnes, James Morris,
Grant Likely, Paul Mackerras, linux-parisc, Magnus Damm,
Jiri Slaby, Dan Williams, devel, Guy Martin, Kuninori Morimoto,
Jeremy Fitzhardinge, Russell King, Thomas Gleixner, Wu
In-Reply-To: <20110120163448.GA30588@suse.de>
On Thu, Jan 20, 2011 at 08:34:48AM -0800, Greg KH wrote:
> On Thu, Jan 20, 2011 at 04:58:13PM +0100, Torben Hohn wrote:
> > the -rt patches change the console_semaphore to console_mutex.
> > so a quite large chunk of the patches changes all
> > acquire/release_console_sem() to acquire/release_console_mutex()
>
> Why not just change the functionality of the existing function to be a
> mutex in the rt patches, instead of having to rename it everywhere?
i hope that Thomas already did this in his upcoming -rt series.
>
> > this commit makes things use more neutral function names
> > which dont make implications about the underlying lock.
> >
> > the only real change is the return value of console_trylock
> > which is inverted from try_acquire_console_sem()
> >
> > Signed-off-by: Torben Hohn <torbenh@gmx.de>
> > CC: Thomas Gleixner <tglx@tglx.de>
>
> I don't mind this rename, but is it really going to help anything out?
> What's the odds of the -rt portion of this patch ever making it to
> mainline?
the -rt portion only changes the semaphore to a mutex.
since the console_sem is used with mutex semantics, i dont see any
reason, not to merge that portion too.
i am just trying to shrink the -rt patch to make it more maintanable :)
>
> thanks,
>
> greg k-h
--
torben Hohn
^ permalink raw reply
* Re: [PATCH] change acquire/release_console_sem() to console_lock/unlock()
From: Greg KH @ 2011-01-20 16:34 UTC (permalink / raw)
To: Torben Hohn
Cc: xen-devel, Kees Cook, linux-fbdev, Sascha Hauer, James Hogan,
David Airlie, Benjamin Herrenschmidt, Jesse Barnes, James Morris,
Grant Likely, Paul Mackerras, linux-parisc, Magnus Damm,
Jiri Slaby, Dan Williams, devel, Guy Martin, Kuninori Morimoto,
Jeremy Fitzhardinge, Russell King, Thomas Gleixner, Wu
In-Reply-To: <1295539104-13210-1-git-send-email-torbenh@gmx.de>
On Thu, Jan 20, 2011 at 04:58:13PM +0100, Torben Hohn wrote:
> the -rt patches change the console_semaphore to console_mutex.
> so a quite large chunk of the patches changes all
> acquire/release_console_sem() to acquire/release_console_mutex()
Why not just change the functionality of the existing function to be a
mutex in the rt patches, instead of having to rename it everywhere?
> this commit makes things use more neutral function names
> which dont make implications about the underlying lock.
>
> the only real change is the return value of console_trylock
> which is inverted from try_acquire_console_sem()
>
> Signed-off-by: Torben Hohn <torbenh@gmx.de>
> CC: Thomas Gleixner <tglx@tglx.de>
I don't mind this rename, but is it really going to help anything out?
What's the odds of the -rt portion of this patch ever making it to
mainline?
thanks,
greg k-h
^ permalink raw reply
* [PATCH] change acquire/release_console_sem() to console_lock/unlock()
From: Torben Hohn @ 2011-01-20 15:58 UTC (permalink / raw)
To: linux-kernel
Cc: xen-devel, Kees Cook, linux-fbdev, Mark Brown, Sascha Hauer,
James Hogan, David Airlie, Benjamin Herrenschmidt, Jesse Barnes,
James Morris, Grant Likely, Paul Mackerras, linux-parisc,
Jiri Slaby, Dan Williams, linux-geode, devel, Guy Martin,
Kuninori Morimoto, Jeremy Fitzhardinge
the -rt patches change the console_semaphore to console_mutex.
so a quite large chunk of the patches changes all
acquire/release_console_sem() to acquire/release_console_mutex()
this commit makes things use more neutral function names
which dont make implications about the underlying lock.
the only real change is the return value of console_trylock
which is inverted from try_acquire_console_sem()
Signed-off-by: Torben Hohn <torbenh@gmx.de>
CC: Thomas Gleixner <tglx@tglx.de>
---
arch/arm/mach-omap2/pm24xx.c | 4 +-
arch/arm/mach-omap2/pm34xx.c | 4 +-
arch/arm/mach-omap2/serial.c | 4 +-
arch/parisc/kernel/pdc_cons.c | 4 +-
drivers/char/bfin_jtag_comm.c | 8 +-
drivers/gpu/drm/nouveau/nouveau_drv.c | 8 +-
drivers/gpu/drm/radeon/radeon_device.c | 10 ++--
drivers/serial/sb1250-duart.c | 2 +-
drivers/staging/msm/msm_fb.c | 8 +-
drivers/staging/olpc_dcon/olpc_dcon.c | 10 ++--
drivers/staging/sm7xx/smtcfb.c | 8 +-
drivers/tty/tty_io.c | 4 +-
drivers/tty/vt/selection.c | 4 +-
drivers/tty/vt/vc_screen.c | 16 ++--
drivers/tty/vt/vt.c | 124 ++++++++++++++++----------------
drivers/tty/vt/vt_ioctl.c | 60 ++++++++--------
drivers/video/arkfb.c | 12 ++--
drivers/video/aty/aty128fb.c | 12 ++--
drivers/video/aty/atyfb_base.c | 10 ++--
drivers/video/aty/radeon_pm.c | 10 ++--
drivers/video/chipsfb.c | 8 +-
drivers/video/console/fbcon.c | 42 ++++++------
drivers/video/da8xx-fb.c | 8 +-
drivers/video/fbmem.c | 12 ++--
drivers/video/fbsysfs.c | 20 +++---
drivers/video/geode/gxfb_core.c | 8 +-
drivers/video/geode/lxfb_core.c | 8 +-
drivers/video/i810/i810_main.c | 8 +-
drivers/video/jz4740_fb.c | 8 +-
drivers/video/mx3fb.c | 8 +-
drivers/video/nvidia/nvidia.c | 8 +-
drivers/video/ps3fb.c | 16 ++--
drivers/video/s3fb.c | 16 ++--
drivers/video/savage/savagefb_driver.c | 8 +-
drivers/video/sh_mobile_hdmi.c | 8 +-
drivers/video/sh_mobile_lcdcfb.c | 4 +-
drivers/video/sm501fb.c | 8 +-
drivers/video/tmiofb.c | 10 ++--
drivers/video/via/viafbdev.c | 8 +-
drivers/video/vt8623fb.c | 12 ++--
drivers/video/xen-fbfront.c | 4 +-
fs/proc/consoles.c | 4 +-
include/linux/console.h | 6 +-
kernel/printk.c | 100 ++++++++++++++------------
44 files changed, 336 insertions(+), 328 deletions(-)
diff --git a/arch/arm/mach-omap2/pm24xx.c b/arch/arm/mach-omap2/pm24xx.c
index 9e5dc8e..97feb3a 100644
--- a/arch/arm/mach-omap2/pm24xx.c
+++ b/arch/arm/mach-omap2/pm24xx.c
@@ -134,7 +134,7 @@ static void omap2_enter_full_retention(void)
/* Block console output in case it is on one of the OMAP UARTs */
if (!is_suspending())
- if (try_acquire_console_sem())
+ if (!console_trylock())
goto no_sleep;
omap_uart_prepare_idle(0);
@@ -151,7 +151,7 @@ static void omap2_enter_full_retention(void)
omap_uart_resume_idle(0);
if (!is_suspending())
- release_console_sem();
+ console_unlock();
no_sleep:
if (omap2_pm_debug) {
diff --git a/arch/arm/mach-omap2/pm34xx.c b/arch/arm/mach-omap2/pm34xx.c
index 8cbbead..a4aa192 100644
--- a/arch/arm/mach-omap2/pm34xx.c
+++ b/arch/arm/mach-omap2/pm34xx.c
@@ -398,7 +398,7 @@ void omap_sram_idle(void)
if (!is_suspending())
if (per_next_state < PWRDM_POWER_ON ||
core_next_state < PWRDM_POWER_ON)
- if (try_acquire_console_sem())
+ if (!console_trylock())
goto console_still_active;
/* PER */
@@ -481,7 +481,7 @@ void omap_sram_idle(void)
}
if (!is_suspending())
- release_console_sem();
+ console_unlock();
console_still_active:
/* Disable IO-PAD and IO-CHAIN wakeup */
diff --git a/arch/arm/mach-omap2/serial.c b/arch/arm/mach-omap2/serial.c
index 302da74..32e91a9 100644
--- a/arch/arm/mach-omap2/serial.c
+++ b/arch/arm/mach-omap2/serial.c
@@ -812,7 +812,7 @@ void __init omap_serial_init_port(struct omap_board_data *bdata)
oh->dev_attr = uart;
- acquire_console_sem(); /* in case the earlycon is on the UART */
+ console_lock(); /* in case the earlycon is on the UART */
/*
* Because of early UART probing, UART did not get idled
@@ -838,7 +838,7 @@ void __init omap_serial_init_port(struct omap_board_data *bdata)
omap_uart_block_sleep(uart);
uart->timeout = DEFAULT_TIMEOUT;
- release_console_sem();
+ console_unlock();
if ((cpu_is_omap34xx() && uart->padconf) ||
(uart->wk_en && uart->wk_mask)) {
diff --git a/arch/parisc/kernel/pdc_cons.c b/arch/parisc/kernel/pdc_cons.c
index 11bdd68..fc770be 100644
--- a/arch/parisc/kernel/pdc_cons.c
+++ b/arch/parisc/kernel/pdc_cons.c
@@ -169,11 +169,11 @@ static int __init pdc_console_tty_driver_init(void)
struct console *tmp;
- acquire_console_sem();
+ console_lock();
for_each_console(tmp)
if (tmp == &pdc_cons)
break;
- release_console_sem();
+ console_unlock();
if (!tmp) {
printk(KERN_INFO "PDC console driver not registered anymore, not creating %s\n", pdc_cons.name);
diff --git a/drivers/char/bfin_jtag_comm.c b/drivers/char/bfin_jtag_comm.c
index e397df3..1640244 100644
--- a/drivers/char/bfin_jtag_comm.c
+++ b/drivers/char/bfin_jtag_comm.c
@@ -183,16 +183,16 @@ bfin_jc_circ_write(const unsigned char *buf, int count)
}
#ifndef CONFIG_BFIN_JTAG_COMM_CONSOLE
-# define acquire_console_sem()
-# define release_console_sem()
+# define console_lock()
+# define console_unlock()
#endif
static int
bfin_jc_write(struct tty_struct *tty, const unsigned char *buf, int count)
{
int i;
- acquire_console_sem();
+ console_lock();
i = bfin_jc_circ_write(buf, count);
- release_console_sem();
+ console_unlock();
wake_up_process(bfin_jc_kthread);
return i;
}
diff --git a/drivers/gpu/drm/nouveau/nouveau_drv.c b/drivers/gpu/drm/nouveau/nouveau_drv.c
index 13bb672..f658a04 100644
--- a/drivers/gpu/drm/nouveau/nouveau_drv.c
+++ b/drivers/gpu/drm/nouveau/nouveau_drv.c
@@ -234,9 +234,9 @@ nouveau_pci_suspend(struct pci_dev *pdev, pm_message_t pm_state)
pci_set_power_state(pdev, PCI_D3hot);
}
- acquire_console_sem();
+ console_lock();
nouveau_fbcon_set_suspend(dev, 1);
- release_console_sem();
+ console_unlock();
nouveau_fbcon_restore_accel(dev);
return 0;
@@ -359,9 +359,9 @@ nouveau_pci_resume(struct pci_dev *pdev)
nv_crtc->lut.depth = 0;
}
- acquire_console_sem();
+ console_lock();
nouveau_fbcon_set_suspend(dev, 0);
- release_console_sem();
+ console_unlock();
nouveau_fbcon_zfill_all(dev);
diff --git a/drivers/gpu/drm/radeon/radeon_device.c b/drivers/gpu/drm/radeon/radeon_device.c
index 26091d6..0d47893 100644
--- a/drivers/gpu/drm/radeon/radeon_device.c
+++ b/drivers/gpu/drm/radeon/radeon_device.c
@@ -891,9 +891,9 @@ int radeon_suspend_kms(struct drm_device *dev, pm_message_t state)
pci_disable_device(dev->pdev);
pci_set_power_state(dev->pdev, PCI_D3hot);
}
- acquire_console_sem();
+ console_lock();
radeon_fbdev_set_suspend(rdev, 1);
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -905,11 +905,11 @@ int radeon_resume_kms(struct drm_device *dev)
if (dev->switch_power_state == DRM_SWITCH_POWER_OFF)
return 0;
- acquire_console_sem();
+ console_lock();
pci_set_power_state(dev->pdev, PCI_D0);
pci_restore_state(dev->pdev);
if (pci_enable_device(dev->pdev)) {
- release_console_sem();
+ console_unlock();
return -1;
}
pci_set_master(dev->pdev);
@@ -920,7 +920,7 @@ int radeon_resume_kms(struct drm_device *dev)
radeon_restore_bios_scratch_regs(rdev);
radeon_fbdev_set_suspend(rdev, 0);
- release_console_sem();
+ console_unlock();
/* reset hpd state */
radeon_hpd_init(rdev);
diff --git a/drivers/serial/sb1250-duart.c b/drivers/serial/sb1250-duart.c
index a2f2b32..602d984 100644
--- a/drivers/serial/sb1250-duart.c
+++ b/drivers/serial/sb1250-duart.c
@@ -829,7 +829,7 @@ static void __init sbd_probe_duarts(void)
#ifdef CONFIG_SERIAL_SB1250_DUART_CONSOLE
/*
* Serial console stuff. Very basic, polling driver for doing serial
- * console output. The console_sem is held by the caller, so we
+ * console output. The console_lock is held by the caller, so we
* shouldn't be interrupted for more console activity.
*/
static void sbd_console_putchar(struct uart_port *uport, int ch)
diff --git a/drivers/staging/msm/msm_fb.c b/drivers/staging/msm/msm_fb.c
index 23fa049..a2f29d4 100644
--- a/drivers/staging/msm/msm_fb.c
+++ b/drivers/staging/msm/msm_fb.c
@@ -347,7 +347,7 @@ static int msm_fb_suspend(struct platform_device *pdev, pm_message_t state)
if ((!mfd) || (mfd->key != MFD_KEY))
return 0;
- acquire_console_sem();
+ console_lock();
fb_set_suspend(mfd->fbi, 1);
ret = msm_fb_suspend_sub(mfd);
@@ -358,7 +358,7 @@ static int msm_fb_suspend(struct platform_device *pdev, pm_message_t state)
pdev->dev.power.power_state = state;
}
- release_console_sem();
+ console_unlock();
return ret;
}
#else
@@ -431,11 +431,11 @@ static int msm_fb_resume(struct platform_device *pdev)
if ((!mfd) || (mfd->key != MFD_KEY))
return 0;
- acquire_console_sem();
+ console_lock();
ret = msm_fb_resume_sub(mfd);
pdev->dev.power.power_state = PMSG_ON;
fb_set_suspend(mfd->fbi, 1);
- release_console_sem();
+ console_unlock();
return ret;
}
diff --git a/drivers/staging/olpc_dcon/olpc_dcon.c b/drivers/staging/olpc_dcon/olpc_dcon.c
index 9f26dc9..56a283d 100644
--- a/drivers/staging/olpc_dcon/olpc_dcon.c
+++ b/drivers/staging/olpc_dcon/olpc_dcon.c
@@ -373,17 +373,17 @@ static void dcon_source_switch(struct work_struct *work)
*
* For now, we just hope..
*/
- acquire_console_sem();
+ console_lock();
ignore_fb_events = 1;
if (fb_blank(fbinfo, FB_BLANK_UNBLANK)) {
ignore_fb_events = 0;
- release_console_sem();
+ console_unlock();
printk(KERN_ERR "olpc-dcon: Failed to enter CPU mode\n");
dcon_pending = DCON_SOURCE_DCON;
return;
}
ignore_fb_events = 0;
- release_console_sem();
+ console_unlock();
/* And turn off the DCON */
pdata->set_dconload(1);
@@ -435,12 +435,12 @@ static void dcon_source_switch(struct work_struct *work)
}
}
- acquire_console_sem();
+ console_lock();
ignore_fb_events = 1;
if (fb_blank(fbinfo, FB_BLANK_POWERDOWN))
printk(KERN_ERR "olpc-dcon: couldn't blank fb!\n");
ignore_fb_events = 0;
- release_console_sem();
+ console_unlock();
printk(KERN_INFO "olpc-dcon: The DCON has control\n");
break;
diff --git a/drivers/staging/sm7xx/smtcfb.c b/drivers/staging/sm7xx/smtcfb.c
index 0bc113c..d007e4a 100644
--- a/drivers/staging/sm7xx/smtcfb.c
+++ b/drivers/staging/sm7xx/smtcfb.c
@@ -1044,9 +1044,9 @@ static int __maybe_unused smtcfb_suspend(struct pci_dev *pdev, pm_message_t msg)
/* when doing suspend, call fb apis and pci apis */
if (msg.event == PM_EVENT_SUSPEND) {
- acquire_console_sem();
+ console_lock();
fb_set_suspend(&sfb->fb, 1);
- release_console_sem();
+ console_unlock();
retv = pci_save_state(pdev);
pci_disable_device(pdev);
retv = pci_choose_state(pdev, msg);
@@ -1105,9 +1105,9 @@ static int __maybe_unused smtcfb_resume(struct pci_dev *pdev)
smtcfb_setmode(sfb);
- acquire_console_sem();
+ console_lock();
fb_set_suspend(&sfb->fb, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/tty/tty_io.c b/drivers/tty/tty_io.c
index 464d09d..6158eae 100644
--- a/drivers/tty/tty_io.c
+++ b/drivers/tty/tty_io.c
@@ -3256,7 +3256,7 @@ static ssize_t show_cons_active(struct device *dev,
struct console *c;
ssize_t count = 0;
- acquire_console_sem();
+ console_lock();
for (c = console_drivers; c; c = c->next) {
if (!c->device)
continue;
@@ -3271,7 +3271,7 @@ static ssize_t show_cons_active(struct device *dev,
while (i--)
count += sprintf(buf + count, "%s%d%c",
cs[i]->name, cs[i]->index, i ? ' ':'\n');
- release_console_sem();
+ console_unlock();
return count;
}
diff --git a/drivers/tty/vt/selection.c b/drivers/tty/vt/selection.c
index ebae344..c956ed6 100644
--- a/drivers/tty/vt/selection.c
+++ b/drivers/tty/vt/selection.c
@@ -316,9 +316,9 @@ int paste_selection(struct tty_struct *tty)
/* always called with BTM from vt_ioctl */
WARN_ON(!tty_locked());
- acquire_console_sem();
+ console_lock();
poke_blanked_console();
- release_console_sem();
+ console_unlock();
ld = tty_ldisc_ref(tty);
if (!ld) {
diff --git a/drivers/tty/vt/vc_screen.c b/drivers/tty/vt/vc_screen.c
index eab3a1f..a672ed1 100644
--- a/drivers/tty/vt/vc_screen.c
+++ b/drivers/tty/vt/vc_screen.c
@@ -202,7 +202,7 @@ vcs_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
/* Select the proper current console and verify
* sanity of the situation under the console lock.
*/
- acquire_console_sem();
+ console_lock();
attr = (currcons & 128);
currcons = (currcons & 127);
@@ -336,9 +336,9 @@ vcs_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
* the pagefault handling code may want to call printk().
*/
- release_console_sem();
+ console_unlock();
ret = copy_to_user(buf, con_buf_start, orig_count);
- acquire_console_sem();
+ console_lock();
if (ret) {
read += (orig_count - ret);
@@ -354,7 +354,7 @@ vcs_read(struct file *file, char __user *buf, size_t count, loff_t *ppos)
if (read)
ret = read;
unlock_out:
- release_console_sem();
+ console_unlock();
mutex_unlock(&con_buf_mtx);
return ret;
}
@@ -379,7 +379,7 @@ vcs_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos)
/* Select the proper current console and verify
* sanity of the situation under the console lock.
*/
- acquire_console_sem();
+ console_lock();
attr = (currcons & 128);
currcons = (currcons & 127);
@@ -414,9 +414,9 @@ vcs_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos)
/* Temporarily drop the console lock so that we can read
* in the write data from userspace safely.
*/
- release_console_sem();
+ console_unlock();
ret = copy_from_user(con_buf, buf, this_round);
- acquire_console_sem();
+ console_lock();
if (ret) {
this_round -= ret;
@@ -542,7 +542,7 @@ vcs_write(struct file *file, const char __user *buf, size_t count, loff_t *ppos)
vcs_scr_updated(vc);
unlock_out:
- release_console_sem();
+ console_unlock();
mutex_unlock(&con_buf_mtx);
diff --git a/drivers/tty/vt/vt.c b/drivers/tty/vt/vt.c
index 76407ec..b230bd3 100644
--- a/drivers/tty/vt/vt.c
+++ b/drivers/tty/vt/vt.c
@@ -1003,9 +1003,9 @@ static int vt_resize(struct tty_struct *tty, struct winsize *ws)
struct vc_data *vc = tty->driver_data;
int ret;
- acquire_console_sem();
+ console_lock();
ret = vc_do_resize(tty, vc, ws->ws_col, ws->ws_row);
- release_console_sem();
+ console_unlock();
return ret;
}
@@ -1271,7 +1271,7 @@ static void default_attr(struct vc_data *vc)
vc->vc_color = vc->vc_def_color;
}
-/* console_sem is held */
+/* console_lock is held */
static void csi_m(struct vc_data *vc)
{
int i;
@@ -1415,7 +1415,7 @@ int mouse_reporting(void)
return vc_cons[fg_console].d->vc_report_mouse;
}
-/* console_sem is held */
+/* console_lock is held */
static void set_mode(struct vc_data *vc, int on_off)
{
int i;
@@ -1485,7 +1485,7 @@ static void set_mode(struct vc_data *vc, int on_off)
}
}
-/* console_sem is held */
+/* console_lock is held */
static void setterm_command(struct vc_data *vc)
{
switch(vc->vc_par[0]) {
@@ -1545,7 +1545,7 @@ static void setterm_command(struct vc_data *vc)
}
}
-/* console_sem is held */
+/* console_lock is held */
static void csi_at(struct vc_data *vc, unsigned int nr)
{
if (nr > vc->vc_cols - vc->vc_x)
@@ -1555,7 +1555,7 @@ static void csi_at(struct vc_data *vc, unsigned int nr)
insert_char(vc, nr);
}
-/* console_sem is held */
+/* console_lock is held */
static void csi_L(struct vc_data *vc, unsigned int nr)
{
if (nr > vc->vc_rows - vc->vc_y)
@@ -1566,7 +1566,7 @@ static void csi_L(struct vc_data *vc, unsigned int nr)
vc->vc_need_wrap = 0;
}
-/* console_sem is held */
+/* console_lock is held */
static void csi_P(struct vc_data *vc, unsigned int nr)
{
if (nr > vc->vc_cols - vc->vc_x)
@@ -1576,7 +1576,7 @@ static void csi_P(struct vc_data *vc, unsigned int nr)
delete_char(vc, nr);
}
-/* console_sem is held */
+/* console_lock is held */
static void csi_M(struct vc_data *vc, unsigned int nr)
{
if (nr > vc->vc_rows - vc->vc_y)
@@ -1587,7 +1587,7 @@ static void csi_M(struct vc_data *vc, unsigned int nr)
vc->vc_need_wrap = 0;
}
-/* console_sem is held (except via vc_init->reset_terminal */
+/* console_lock is held (except via vc_init->reset_terminal */
static void save_cur(struct vc_data *vc)
{
vc->vc_saved_x = vc->vc_x;
@@ -1603,7 +1603,7 @@ static void save_cur(struct vc_data *vc)
vc->vc_saved_G1 = vc->vc_G1_charset;
}
-/* console_sem is held */
+/* console_lock is held */
static void restore_cur(struct vc_data *vc)
{
gotoxy(vc, vc->vc_saved_x, vc->vc_saved_y);
@@ -1625,7 +1625,7 @@ enum { ESnormal, ESesc, ESsquare, ESgetpars, ESgotpars, ESfunckey,
EShash, ESsetG0, ESsetG1, ESpercent, ESignore, ESnonstd,
ESpalette };
-/* console_sem is held (except via vc_init()) */
+/* console_lock is held (except via vc_init()) */
static void reset_terminal(struct vc_data *vc, int do_clear)
{
vc->vc_top = 0;
@@ -1685,7 +1685,7 @@ static void reset_terminal(struct vc_data *vc, int do_clear)
csi_J(vc, 2);
}
-/* console_sem is held */
+/* console_lock is held */
static void do_con_trol(struct tty_struct *tty, struct vc_data *vc, int c)
{
/*
@@ -2119,7 +2119,7 @@ static int is_double_width(uint32_t ucs)
return bisearch(ucs, double_width, ARRAY_SIZE(double_width) - 1);
}
-/* acquires console_sem */
+/* acquires console_lock */
static int do_con_write(struct tty_struct *tty, const unsigned char *buf, int count)
{
#ifdef VT_BUF_VRAM_ONLY
@@ -2147,11 +2147,11 @@ static int do_con_write(struct tty_struct *tty, const unsigned char *buf, int co
might_sleep();
- acquire_console_sem();
+ console_lock();
vc = tty->driver_data;
if (vc == NULL) {
printk(KERN_ERR "vt: argh, driver_data is NULL !\n");
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -2159,7 +2159,7 @@ static int do_con_write(struct tty_struct *tty, const unsigned char *buf, int co
if (!vc_cons_allocated(currcons)) {
/* could this happen? */
printk_once("con_write: tty %d not allocated\n", currcons+1);
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -2375,7 +2375,7 @@ rescan_last_byte:
}
FLUSH
console_conditional_schedule();
- release_console_sem();
+ console_unlock();
notify_update(vc);
return n;
#undef FLUSH
@@ -2388,11 +2388,11 @@ rescan_last_byte:
* us to do the switches asynchronously (needed when we want
* to switch due to a keyboard interrupt). Synchronization
* with other console code and prevention of re-entrancy is
- * ensured with console_sem.
+ * ensured with console_lock.
*/
static void console_callback(struct work_struct *ignored)
{
- acquire_console_sem();
+ console_lock();
if (want_console >= 0) {
if (want_console != fg_console &&
@@ -2422,7 +2422,7 @@ static void console_callback(struct work_struct *ignored)
}
notify_update(vc_cons[fg_console].d);
- release_console_sem();
+ console_unlock();
}
int set_console(int nr)
@@ -2603,7 +2603,7 @@ static struct console vt_console_driver = {
*/
/*
- * Generally a bit racy with respect to console_sem().
+ * Generally a bit racy with respect to console_lock();.
*
* There are some functions which don't need it.
*
@@ -2629,17 +2629,17 @@ int tioclinux(struct tty_struct *tty, unsigned long arg)
switch (type)
{
case TIOCL_SETSEL:
- acquire_console_sem();
+ console_lock();
ret = set_selection((struct tiocl_selection __user *)(p+1), tty);
- release_console_sem();
+ console_unlock();
break;
case TIOCL_PASTESEL:
ret = paste_selection(tty);
break;
case TIOCL_UNBLANKSCREEN:
- acquire_console_sem();
+ console_lock();
unblank_screen();
- release_console_sem();
+ console_unlock();
break;
case TIOCL_SELLOADLUT:
ret = sel_loadlut(p);
@@ -2688,10 +2688,10 @@ int tioclinux(struct tty_struct *tty, unsigned long arg)
}
break;
case TIOCL_BLANKSCREEN: /* until explicitly unblanked, not only poked */
- acquire_console_sem();
+ console_lock();
ignore_poke = 1;
do_blank_screen(0);
- release_console_sem();
+ console_unlock();
break;
case TIOCL_BLANKEDSCREEN:
ret = console_blanked;
@@ -2790,11 +2790,11 @@ static void con_flush_chars(struct tty_struct *tty)
return;
/* if we race with con_close(), vt may be null */
- acquire_console_sem();
+ console_lock();
vc = tty->driver_data;
if (vc)
set_cursor(vc);
- release_console_sem();
+ console_unlock();
}
/*
@@ -2805,7 +2805,7 @@ static int con_open(struct tty_struct *tty, struct file *filp)
unsigned int currcons = tty->index;
int ret = 0;
- acquire_console_sem();
+ console_lock();
if (tty->driver_data == NULL) {
ret = vc_allocate(currcons);
if (ret == 0) {
@@ -2813,7 +2813,7 @@ static int con_open(struct tty_struct *tty, struct file *filp)
/* Still being freed */
if (vc->port.tty) {
- release_console_sem();
+ console_unlock();
return -ERESTARTSYS;
}
tty->driver_data = vc;
@@ -2827,11 +2827,11 @@ static int con_open(struct tty_struct *tty, struct file *filp)
tty->termios->c_iflag |= IUTF8;
else
tty->termios->c_iflag &= ~IUTF8;
- release_console_sem();
+ console_unlock();
return ret;
}
}
- release_console_sem();
+ console_unlock();
return ret;
}
@@ -2844,9 +2844,9 @@ static void con_shutdown(struct tty_struct *tty)
{
struct vc_data *vc = tty->driver_data;
BUG_ON(vc == NULL);
- acquire_console_sem();
+ console_lock();
vc->port.tty = NULL;
- release_console_sem();
+ console_unlock();
tty_shutdown(tty);
}
@@ -2893,13 +2893,13 @@ static int __init con_init(void)
struct vc_data *vc;
unsigned int currcons = 0, i;
- acquire_console_sem();
+ console_lock();
if (conswitchp)
display_desc = conswitchp->con_startup();
if (!display_desc) {
fg_console = 0;
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -2946,7 +2946,7 @@ static int __init con_init(void)
printable = 1;
printk("\n");
- release_console_sem();
+ console_unlock();
#ifdef CONFIG_VT_CONSOLE
register_console(&vt_console_driver);
@@ -3037,7 +3037,7 @@ static int bind_con_driver(const struct consw *csw, int first, int last,
if (!try_module_get(owner))
return -ENODEV;
- acquire_console_sem();
+ console_lock();
/* check if driver is registered */
for (i = 0; i < MAX_NR_CON_DRIVER; i++) {
@@ -3122,7 +3122,7 @@ static int bind_con_driver(const struct consw *csw, int first, int last,
retval = 0;
err:
- release_console_sem();
+ console_unlock();
module_put(owner);
return retval;
};
@@ -3171,7 +3171,7 @@ int unbind_con_driver(const struct consw *csw, int first, int last, int deflt)
if (!try_module_get(owner))
return -ENODEV;
- acquire_console_sem();
+ console_lock();
/* check if driver is registered and if it is unbindable */
for (i = 0; i < MAX_NR_CON_DRIVER; i++) {
@@ -3185,7 +3185,7 @@ int unbind_con_driver(const struct consw *csw, int first, int last, int deflt)
}
if (retval) {
- release_console_sem();
+ console_unlock();
goto err;
}
@@ -3204,12 +3204,12 @@ int unbind_con_driver(const struct consw *csw, int first, int last, int deflt)
}
if (retval) {
- release_console_sem();
+ console_unlock();
goto err;
}
if (!con_is_bound(csw)) {
- release_console_sem();
+ console_unlock();
goto err;
}
@@ -3238,7 +3238,7 @@ int unbind_con_driver(const struct consw *csw, int first, int last, int deflt)
if (!con_is_bound(csw))
con_driver->flag &= ~CON_DRIVER_FLAG_INIT;
- release_console_sem();
+ console_unlock();
/* ignore return value, binding should not fail */
bind_con_driver(defcsw, first, last, deflt);
err:
@@ -3538,7 +3538,7 @@ int register_con_driver(const struct consw *csw, int first, int last)
if (!try_module_get(owner))
return -ENODEV;
- acquire_console_sem();
+ console_lock();
for (i = 0; i < MAX_NR_CON_DRIVER; i++) {
con_driver = ®istered_con_driver[i];
@@ -3592,7 +3592,7 @@ int register_con_driver(const struct consw *csw, int first, int last)
}
err:
- release_console_sem();
+ console_unlock();
module_put(owner);
return retval;
}
@@ -3613,7 +3613,7 @@ int unregister_con_driver(const struct consw *csw)
{
int i, retval = -ENODEV;
- acquire_console_sem();
+ console_lock();
/* cannot unregister a bound driver */
if (con_is_bound(csw))
@@ -3639,7 +3639,7 @@ int unregister_con_driver(const struct consw *csw)
}
}
err:
- release_console_sem();
+ console_unlock();
return retval;
}
EXPORT_SYMBOL(unregister_con_driver);
@@ -3934,9 +3934,9 @@ int con_set_cmap(unsigned char __user *arg)
{
int rc;
- acquire_console_sem();
+ console_lock();
rc = set_get_cmap (arg,1);
- release_console_sem();
+ console_unlock();
return rc;
}
@@ -3945,9 +3945,9 @@ int con_get_cmap(unsigned char __user *arg)
{
int rc;
- acquire_console_sem();
+ console_lock();
rc = set_get_cmap (arg,0);
- release_console_sem();
+ console_unlock();
return rc;
}
@@ -3994,12 +3994,12 @@ static int con_font_get(struct vc_data *vc, struct console_font_op *op)
} else
font.data = NULL;
- acquire_console_sem();
+ console_lock();
if (vc->vc_sw->con_font_get)
rc = vc->vc_sw->con_font_get(vc, &font);
else
rc = -ENOSYS;
- release_console_sem();
+ console_unlock();
if (rc)
goto out;
@@ -4076,12 +4076,12 @@ static int con_font_set(struct vc_data *vc, struct console_font_op *op)
font.data = memdup_user(op->data, size);
if (IS_ERR(font.data))
return PTR_ERR(font.data);
- acquire_console_sem();
+ console_lock();
if (vc->vc_sw->con_font_set)
rc = vc->vc_sw->con_font_set(vc, &font, op->flags);
else
rc = -ENOSYS;
- release_console_sem();
+ console_unlock();
kfree(font.data);
return rc;
}
@@ -4103,12 +4103,12 @@ static int con_font_default(struct vc_data *vc, struct console_font_op *op)
else
name[MAX_FONT_NAME - 1] = 0;
- acquire_console_sem();
+ console_lock();
if (vc->vc_sw->con_font_default)
rc = vc->vc_sw->con_font_default(vc, &font, s);
else
rc = -ENOSYS;
- release_console_sem();
+ console_unlock();
if (!rc) {
op->width = font.width;
op->height = font.height;
@@ -4124,7 +4124,7 @@ static int con_font_copy(struct vc_data *vc, struct console_font_op *op)
if (vc->vc_mode != KD_TEXT)
return -EINVAL;
- acquire_console_sem();
+ console_lock();
if (!vc->vc_sw->con_font_copy)
rc = -ENOSYS;
else if (con < 0 || !vc_cons_allocated(con))
@@ -4133,7 +4133,7 @@ static int con_font_copy(struct vc_data *vc, struct console_font_op *op)
rc = 0;
else
rc = vc->vc_sw->con_font_copy(vc, con);
- release_console_sem();
+ console_unlock();
return rc;
}
diff --git a/drivers/tty/vt/vt_ioctl.c b/drivers/tty/vt/vt_ioctl.c
index 6b68a0f..1235ebd 100644
--- a/drivers/tty/vt/vt_ioctl.c
+++ b/drivers/tty/vt/vt_ioctl.c
@@ -649,12 +649,12 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
/*
* explicitly blank/unblank the screen if switching modes
*/
- acquire_console_sem();
+ console_lock();
if (arg == KD_TEXT)
do_unblank_screen(1);
else
do_blank_screen(1);
- release_console_sem();
+ console_unlock();
break;
case KDGETMODE:
@@ -893,7 +893,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
ret = -EINVAL;
goto out;
}
- acquire_console_sem();
+ console_lock();
vc->vt_mode = tmp;
/* the frsig is ignored, so we set it to 0 */
vc->vt_mode.frsig = 0;
@@ -901,7 +901,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
vc->vt_pid = get_pid(task_pid(current));
/* no switch is required -- saw@shade.msu.ru */
vc->vt_newvt = -1;
- release_console_sem();
+ console_unlock();
break;
}
@@ -910,9 +910,9 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
struct vt_mode tmp;
int rc;
- acquire_console_sem();
+ console_lock();
memcpy(&tmp, &vc->vt_mode, sizeof(struct vt_mode));
- release_console_sem();
+ console_unlock();
rc = copy_to_user(up, &tmp, sizeof(struct vt_mode));
if (rc)
@@ -965,9 +965,9 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
ret = -ENXIO;
else {
arg--;
- acquire_console_sem();
+ console_lock();
ret = vc_allocate(arg);
- release_console_sem();
+ console_unlock();
if (ret)
break;
set_console(arg);
@@ -990,7 +990,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
ret = -ENXIO;
else {
vsa.console--;
- acquire_console_sem();
+ console_lock();
ret = vc_allocate(vsa.console);
if (ret == 0) {
struct vc_data *nvc;
@@ -1003,7 +1003,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
put_pid(nvc->vt_pid);
nvc->vt_pid = get_pid(task_pid(current));
}
- release_console_sem();
+ console_unlock();
if (ret)
break;
/* Commence switch and lock */
@@ -1044,7 +1044,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
/*
* Switching-from response
*/
- acquire_console_sem();
+ console_lock();
if (vc->vt_newvt >= 0) {
if (arg == 0)
/*
@@ -1063,7 +1063,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
vc->vt_newvt = -1;
ret = vc_allocate(newvt);
if (ret) {
- release_console_sem();
+ console_unlock();
break;
}
/*
@@ -1083,7 +1083,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
if (arg != VT_ACKACQ)
ret = -EINVAL;
}
- release_console_sem();
+ console_unlock();
break;
/*
@@ -1096,20 +1096,20 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
}
if (arg == 0) {
/* deallocate all unused consoles, but leave 0 */
- acquire_console_sem();
+ console_lock();
for (i=1; i<MAX_NR_CONSOLES; i++)
if (! VT_BUSY(i))
vc_deallocate(i);
- release_console_sem();
+ console_unlock();
} else {
/* deallocate a single console, if possible */
arg--;
if (VT_BUSY(arg))
ret = -EBUSY;
else if (arg) { /* leave 0 */
- acquire_console_sem();
+ console_lock();
vc_deallocate(arg);
- release_console_sem();
+ console_unlock();
}
}
break;
@@ -1126,7 +1126,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
get_user(cc, &vtsizes->v_cols))
ret = -EFAULT;
else {
- acquire_console_sem();
+ console_lock();
for (i = 0; i < MAX_NR_CONSOLES; i++) {
vc = vc_cons[i].d;
@@ -1135,7 +1135,7 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
vc_resize(vc_cons[i].d, cc, ll);
}
}
- release_console_sem();
+ console_unlock();
}
break;
}
@@ -1187,14 +1187,14 @@ int vt_ioctl(struct tty_struct *tty, struct file * file,
for (i = 0; i < MAX_NR_CONSOLES; i++) {
if (!vc_cons[i].d)
continue;
- acquire_console_sem();
+ console_lock();
if (vlin)
vc_cons[i].d->vc_scan_lines = vlin;
if (clin)
vc_cons[i].d->vc_font.height = clin;
vc_cons[i].d->vc_resize_user = 1;
vc_resize(vc_cons[i].d, cc, ll);
- release_console_sem();
+ console_unlock();
}
break;
}
@@ -1367,7 +1367,7 @@ void vc_SAK(struct work_struct *work)
struct vc_data *vc;
struct tty_struct *tty;
- acquire_console_sem();
+ console_lock();
vc = vc_con->d;
if (vc) {
tty = vc->port.tty;
@@ -1379,7 +1379,7 @@ void vc_SAK(struct work_struct *work)
__do_SAK(tty);
reset_vc(vc);
}
- release_console_sem();
+ console_unlock();
}
#ifdef CONFIG_COMPAT
@@ -1737,10 +1737,10 @@ int vt_move_to_console(unsigned int vt, int alloc)
{
int prev;
- acquire_console_sem();
+ console_lock();
/* Graphics mode - up to X */
if (disable_vt_switch) {
- release_console_sem();
+ console_unlock();
return 0;
}
prev = fg_console;
@@ -1748,7 +1748,7 @@ int vt_move_to_console(unsigned int vt, int alloc)
if (alloc && vc_allocate(vt)) {
/* we can't have a free VC for now. Too bad,
* we don't want to mess the screen for now. */
- release_console_sem();
+ console_unlock();
return -ENOSPC;
}
@@ -1758,10 +1758,10 @@ int vt_move_to_console(unsigned int vt, int alloc)
* Let the calling function know so it can decide
* what to do.
*/
- release_console_sem();
+ console_unlock();
return -EIO;
}
- release_console_sem();
+ console_unlock();
tty_lock();
if (vt_waitactive(vt + 1)) {
pr_debug("Suspend: Can't switch VCs.");
@@ -1781,8 +1781,8 @@ int vt_move_to_console(unsigned int vt, int alloc)
*/
void pm_set_vt_switch(int do_switch)
{
- acquire_console_sem();
+ console_lock();
disable_vt_switch = !do_switch;
- release_console_sem();
+ console_unlock();
}
EXPORT_SYMBOL(pm_set_vt_switch);
diff --git a/drivers/video/arkfb.c b/drivers/video/arkfb.c
index d583bea..391ac93 100644
--- a/drivers/video/arkfb.c
+++ b/drivers/video/arkfb.c
@@ -23,7 +23,7 @@
#include <linux/svga.h>
#include <linux/init.h>
#include <linux/pci.h>
-#include <linux/console.h> /* Why should fb driver call console functions? because acquire_console_sem() */
+#include <linux/console.h> /* Why should fb driver call console functions? because console_lock() */
#include <video/vga.h>
#ifdef CONFIG_MTRR
@@ -1091,12 +1091,12 @@ static int ark_pci_suspend (struct pci_dev* dev, pm_message_t state)
dev_info(info->device, "suspend\n");
- acquire_console_sem();
+ console_lock();
mutex_lock(&(par->open_lock));
if ((state.event == PM_EVENT_FREEZE) || (par->ref_count == 0)) {
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1107,7 +1107,7 @@ static int ark_pci_suspend (struct pci_dev* dev, pm_message_t state)
pci_set_power_state(dev, pci_choose_state(dev, state));
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1122,7 +1122,7 @@ static int ark_pci_resume (struct pci_dev* dev)
dev_info(info->device, "resume\n");
- acquire_console_sem();
+ console_lock();
mutex_lock(&(par->open_lock));
if (par->ref_count == 0)
@@ -1141,7 +1141,7 @@ static int ark_pci_resume (struct pci_dev* dev)
fail:
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
#else
diff --git a/drivers/video/aty/aty128fb.c b/drivers/video/aty/aty128fb.c
index dd9de2e..4cb6a57 100644
--- a/drivers/video/aty/aty128fb.c
+++ b/drivers/video/aty/aty128fb.c
@@ -1860,11 +1860,11 @@ static void aty128_early_resume(void *data)
{
struct aty128fb_par *par = data;
- if (try_acquire_console_sem())
+ if (!console_trylock())
return;
pci_restore_state(par->pdev);
aty128_do_resume(par->pdev);
- release_console_sem();
+ console_unlock();
}
#endif /* CONFIG_PPC_PMAC */
@@ -2438,7 +2438,7 @@ static int aty128_pci_suspend(struct pci_dev *pdev, pm_message_t state)
printk(KERN_DEBUG "aty128fb: suspending...\n");
- acquire_console_sem();
+ console_lock();
fb_set_suspend(info, 1);
@@ -2470,7 +2470,7 @@ static int aty128_pci_suspend(struct pci_dev *pdev, pm_message_t state)
if (state.event != PM_EVENT_ON)
aty128_set_suspend(par, 1);
- release_console_sem();
+ console_unlock();
pdev->dev.power.power_state = state;
@@ -2527,9 +2527,9 @@ static int aty128_pci_resume(struct pci_dev *pdev)
{
int rc;
- acquire_console_sem();
+ console_lock();
rc = aty128_do_resume(pdev);
- release_console_sem();
+ console_unlock();
return rc;
}
diff --git a/drivers/video/aty/atyfb_base.c b/drivers/video/aty/atyfb_base.c
index 767ab4f..94e293f 100644
--- a/drivers/video/aty/atyfb_base.c
+++ b/drivers/video/aty/atyfb_base.c
@@ -2069,7 +2069,7 @@ static int atyfb_pci_suspend(struct pci_dev *pdev, pm_message_t state)
if (state.event == pdev->dev.power.power_state.event)
return 0;
- acquire_console_sem();
+ console_lock();
fb_set_suspend(info, 1);
@@ -2097,14 +2097,14 @@ static int atyfb_pci_suspend(struct pci_dev *pdev, pm_message_t state)
par->lock_blank = 0;
atyfb_blank(FB_BLANK_UNBLANK, info);
fb_set_suspend(info, 0);
- release_console_sem();
+ console_unlock();
return -EIO;
}
#else
pci_set_power_state(pdev, pci_choose_state(pdev, state));
#endif
- release_console_sem();
+ console_unlock();
pdev->dev.power.power_state = state;
@@ -2133,7 +2133,7 @@ static int atyfb_pci_resume(struct pci_dev *pdev)
if (pdev->dev.power.power_state.event == PM_EVENT_ON)
return 0;
- acquire_console_sem();
+ console_lock();
/*
* PCI state will have been restored by the core, so
@@ -2161,7 +2161,7 @@ static int atyfb_pci_resume(struct pci_dev *pdev)
par->lock_blank = 0;
atyfb_blank(FB_BLANK_UNBLANK, info);
- release_console_sem();
+ console_unlock();
pdev->dev.power.power_state = PMSG_ON;
diff --git a/drivers/video/aty/radeon_pm.c b/drivers/video/aty/radeon_pm.c
index c4e1764..92bda58 100644
--- a/drivers/video/aty/radeon_pm.c
+++ b/drivers/video/aty/radeon_pm.c
@@ -2626,7 +2626,7 @@ int radeonfb_pci_suspend(struct pci_dev *pdev, pm_message_t mesg)
goto done;
}
- acquire_console_sem();
+ console_lock();
fb_set_suspend(info, 1);
@@ -2690,7 +2690,7 @@ int radeonfb_pci_suspend(struct pci_dev *pdev, pm_message_t mesg)
if (rinfo->pm_mode & radeon_pm_d2)
radeon_set_suspend(rinfo, 1);
- release_console_sem();
+ console_unlock();
done:
pdev->dev.power.power_state = mesg;
@@ -2715,10 +2715,10 @@ int radeonfb_pci_resume(struct pci_dev *pdev)
return 0;
if (rinfo->no_schedule) {
- if (try_acquire_console_sem())
+ if (!console_trylock())
return 0;
} else
- acquire_console_sem();
+ console_lock();
printk(KERN_DEBUG "radeonfb (%s): resuming from state: %d...\n",
pci_name(pdev), pdev->dev.power.power_state.event);
@@ -2783,7 +2783,7 @@ int radeonfb_pci_resume(struct pci_dev *pdev)
pdev->dev.power.power_state = PMSG_ON;
bail:
- release_console_sem();
+ console_unlock();
return rc;
}
diff --git a/drivers/video/chipsfb.c b/drivers/video/chipsfb.c
index d637e1f..cff742a 100644
--- a/drivers/video/chipsfb.c
+++ b/drivers/video/chipsfb.c
@@ -460,10 +460,10 @@ static int chipsfb_pci_suspend(struct pci_dev *pdev, pm_message_t state)
if (!(state.event & PM_EVENT_SLEEP))
goto done;
- acquire_console_sem();
+ console_lock();
chipsfb_blank(1, p);
fb_set_suspend(p, 1);
- release_console_sem();
+ console_unlock();
done:
pdev->dev.power.power_state = state;
return 0;
@@ -473,10 +473,10 @@ static int chipsfb_pci_resume(struct pci_dev *pdev)
{
struct fb_info *p = pci_get_drvdata(pdev);
- acquire_console_sem();
+ console_lock();
fb_set_suspend(p, 0);
chipsfb_blank(0, p);
- release_console_sem();
+ console_unlock();
pdev->dev.power.power_state = PMSG_ON;
return 0;
diff --git a/drivers/video/console/fbcon.c b/drivers/video/console/fbcon.c
index 7ccc967..9c092b8 100644
--- a/drivers/video/console/fbcon.c
+++ b/drivers/video/console/fbcon.c
@@ -375,14 +375,14 @@ static void fb_flashcursor(struct work_struct *work)
int c;
int mode;
- acquire_console_sem();
+ console_lock();
if (ops && ops->currcon != -1)
vc = vc_cons[ops->currcon].d;
if (!vc || !CON_IS_VISIBLE(vc) ||
registered_fb[con2fb_map[vc->vc_num]] != info ||
vc->vc_deccm != 1) {
- release_console_sem();
+ console_unlock();
return;
}
@@ -392,7 +392,7 @@ static void fb_flashcursor(struct work_struct *work)
CM_ERASE : CM_DRAW;
ops->cursor(vc, info, mode, softback_lines, get_color(vc, info, c, 1),
get_color(vc, info, c, 0));
- release_console_sem();
+ console_unlock();
}
static void cursor_timer_handler(unsigned long dev_addr)
@@ -836,7 +836,7 @@ static int set_con2fb_map(int unit, int newidx, int user)
found = search_fb_in_map(newidx);
- acquire_console_sem();
+ console_lock();
con2fb_map[unit] = newidx;
if (!err && !found)
err = con2fb_acquire_newinfo(vc, info, unit, oldidx);
@@ -863,7 +863,7 @@ static int set_con2fb_map(int unit, int newidx, int user)
if (!search_fb_in_map(info_idx))
info_idx = newidx;
- release_console_sem();
+ console_unlock();
return err;
}
@@ -3321,7 +3321,7 @@ static ssize_t store_rotate(struct device *device,
if (fbcon_has_exited)
return count;
- acquire_console_sem();
+ console_lock();
idx = con2fb_map[fg_console];
if (idx == -1 || registered_fb[idx] == NULL)
@@ -3331,7 +3331,7 @@ static ssize_t store_rotate(struct device *device,
rotate = simple_strtoul(buf, last, 0);
fbcon_rotate(info, rotate);
err:
- release_console_sem();
+ console_unlock();
return count;
}
@@ -3346,7 +3346,7 @@ static ssize_t store_rotate_all(struct device *device,
if (fbcon_has_exited)
return count;
- acquire_console_sem();
+ console_lock();
idx = con2fb_map[fg_console];
if (idx == -1 || registered_fb[idx] == NULL)
@@ -3356,7 +3356,7 @@ static ssize_t store_rotate_all(struct device *device,
rotate = simple_strtoul(buf, last, 0);
fbcon_rotate_all(info, rotate);
err:
- release_console_sem();
+ console_unlock();
return count;
}
@@ -3369,7 +3369,7 @@ static ssize_t show_rotate(struct device *device,
if (fbcon_has_exited)
return 0;
- acquire_console_sem();
+ console_lock();
idx = con2fb_map[fg_console];
if (idx == -1 || registered_fb[idx] == NULL)
@@ -3378,7 +3378,7 @@ static ssize_t show_rotate(struct device *device,
info = registered_fb[idx];
rotate = fbcon_get_rotate(info);
err:
- release_console_sem();
+ console_unlock();
return snprintf(buf, PAGE_SIZE, "%d\n", rotate);
}
@@ -3392,7 +3392,7 @@ static ssize_t show_cursor_blink(struct device *device,
if (fbcon_has_exited)
return 0;
- acquire_console_sem();
+ console_lock();
idx = con2fb_map[fg_console];
if (idx == -1 || registered_fb[idx] == NULL)
@@ -3406,7 +3406,7 @@ static ssize_t show_cursor_blink(struct device *device,
blink = (ops->flags & FBCON_FLAGS_CURSOR_TIMER) ? 1 : 0;
err:
- release_console_sem();
+ console_unlock();
return snprintf(buf, PAGE_SIZE, "%d\n", blink);
}
@@ -3421,7 +3421,7 @@ static ssize_t store_cursor_blink(struct device *device,
if (fbcon_has_exited)
return count;
- acquire_console_sem();
+ console_lock();
idx = con2fb_map[fg_console];
if (idx == -1 || registered_fb[idx] == NULL)
@@ -3443,7 +3443,7 @@ static ssize_t store_cursor_blink(struct device *device,
}
err:
- release_console_sem();
+ console_unlock();
return count;
}
@@ -3482,7 +3482,7 @@ static void fbcon_start(void)
if (num_registered_fb) {
int i;
- acquire_console_sem();
+ console_lock();
for (i = 0; i < FB_MAX; i++) {
if (registered_fb[i] != NULL) {
@@ -3491,7 +3491,7 @@ static void fbcon_start(void)
}
}
- release_console_sem();
+ console_unlock();
fbcon_takeover(0);
}
}
@@ -3552,7 +3552,7 @@ static int __init fb_console_init(void)
{
int i;
- acquire_console_sem();
+ console_lock();
fb_register_client(&fbcon_event_notifier);
fbcon_device = device_create(fb_class, NULL, MKDEV(0, 0), NULL,
"fbcon");
@@ -3568,7 +3568,7 @@ static int __init fb_console_init(void)
for (i = 0; i < MAX_NR_CONSOLES; i++)
con2fb_map[i] = -1;
- release_console_sem();
+ console_unlock();
fbcon_start();
return 0;
}
@@ -3591,12 +3591,12 @@ static void __exit fbcon_deinit_device(void)
static void __exit fb_console_exit(void)
{
- acquire_console_sem();
+ console_lock();
fb_unregister_client(&fbcon_event_notifier);
fbcon_deinit_device();
device_destroy(fb_class, MKDEV(0, 0));
fbcon_exit();
- release_console_sem();
+ console_unlock();
unregister_con_driver(&fb_con);
}
diff --git a/drivers/video/da8xx-fb.c b/drivers/video/da8xx-fb.c
index c265aed..fbf6a9c 100644
--- a/drivers/video/da8xx-fb.c
+++ b/drivers/video/da8xx-fb.c
@@ -1130,14 +1130,14 @@ static int fb_suspend(struct platform_device *dev, pm_message_t state)
struct fb_info *info = platform_get_drvdata(dev);
struct da8xx_fb_par *par = info->par;
- acquire_console_sem();
+ console_lock();
if (par->panel_power_ctrl)
par->panel_power_ctrl(0);
fb_set_suspend(info, 1);
lcd_disable_raster();
clk_disable(par->lcdc_clk);
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1146,14 +1146,14 @@ static int fb_resume(struct platform_device *dev)
struct fb_info *info = platform_get_drvdata(dev);
struct da8xx_fb_par *par = info->par;
- acquire_console_sem();
+ console_lock();
if (par->panel_power_ctrl)
par->panel_power_ctrl(1);
clk_enable(par->lcdc_clk);
lcd_enable_raster();
fb_set_suspend(info, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/fbmem.c b/drivers/video/fbmem.c
index 4ac1201..e2bf953 100644
--- a/drivers/video/fbmem.c
+++ b/drivers/video/fbmem.c
@@ -1036,11 +1036,11 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
return -EFAULT;
if (!lock_fb_info(info))
return -ENODEV;
- acquire_console_sem();
+ console_lock();
info->flags |= FBINFO_MISC_USEREVENT;
ret = fb_set_var(info, &var);
info->flags &= ~FBINFO_MISC_USEREVENT;
- release_console_sem();
+ console_unlock();
unlock_fb_info(info);
if (!ret && copy_to_user(argp, &var, sizeof(var)))
ret = -EFAULT;
@@ -1072,9 +1072,9 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
return -EFAULT;
if (!lock_fb_info(info))
return -ENODEV;
- acquire_console_sem();
+ console_lock();
ret = fb_pan_display(info, &var);
- release_console_sem();
+ console_unlock();
unlock_fb_info(info);
if (ret == 0 && copy_to_user(argp, &var, sizeof(var)))
return -EFAULT;
@@ -1119,11 +1119,11 @@ static long do_fb_ioctl(struct fb_info *info, unsigned int cmd,
case FBIOBLANK:
if (!lock_fb_info(info))
return -ENODEV;
- acquire_console_sem();
+ console_lock();
info->flags |= FBINFO_MISC_USEREVENT;
ret = fb_blank(info, arg);
info->flags &= ~FBINFO_MISC_USEREVENT;
- release_console_sem();
+ console_unlock();
unlock_fb_info(info);
break;
default:
diff --git a/drivers/video/fbsysfs.c b/drivers/video/fbsysfs.c
index 0a08f13..f4a3277 100644
--- a/drivers/video/fbsysfs.c
+++ b/drivers/video/fbsysfs.c
@@ -90,11 +90,11 @@ static int activate(struct fb_info *fb_info, struct fb_var_screeninfo *var)
int err;
var->activate |= FB_ACTIVATE_FORCE;
- acquire_console_sem();
+ console_lock();
fb_info->flags |= FBINFO_MISC_USEREVENT;
err = fb_set_var(fb_info, var);
fb_info->flags &= ~FBINFO_MISC_USEREVENT;
- release_console_sem();
+ console_unlock();
if (err)
return err;
return 0;
@@ -175,7 +175,7 @@ static ssize_t store_modes(struct device *device,
if (i * sizeof(struct fb_videomode) != count)
return -EINVAL;
- acquire_console_sem();
+ console_lock();
list_splice(&fb_info->modelist, &old_list);
fb_videomode_to_modelist((const struct fb_videomode *)buf, i,
&fb_info->modelist);
@@ -185,7 +185,7 @@ static ssize_t store_modes(struct device *device,
} else
fb_destroy_modelist(&old_list);
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -301,11 +301,11 @@ static ssize_t store_blank(struct device *device,
char *last = NULL;
int err;
- acquire_console_sem();
+ console_lock();
fb_info->flags |= FBINFO_MISC_USEREVENT;
err = fb_blank(fb_info, simple_strtoul(buf, &last, 0));
fb_info->flags &= ~FBINFO_MISC_USEREVENT;
- release_console_sem();
+ console_unlock();
if (err < 0)
return err;
return count;
@@ -364,9 +364,9 @@ static ssize_t store_pan(struct device *device,
return -EINVAL;
var.yoffset = simple_strtoul(last, &last, 0);
- acquire_console_sem();
+ console_lock();
err = fb_pan_display(fb_info, &var);
- release_console_sem();
+ console_unlock();
if (err < 0)
return err;
@@ -399,9 +399,9 @@ static ssize_t store_fbstate(struct device *device,
state = simple_strtoul(buf, &last, 0);
- acquire_console_sem();
+ console_lock();
fb_set_suspend(fb_info, (int)state);
- release_console_sem();
+ console_unlock();
return count;
}
diff --git a/drivers/video/geode/gxfb_core.c b/drivers/video/geode/gxfb_core.c
index 70b1d9d..b4f19db 100644
--- a/drivers/video/geode/gxfb_core.c
+++ b/drivers/video/geode/gxfb_core.c
@@ -344,10 +344,10 @@ static int gxfb_suspend(struct pci_dev *pdev, pm_message_t state)
struct fb_info *info = pci_get_drvdata(pdev);
if (state.event == PM_EVENT_SUSPEND) {
- acquire_console_sem();
+ console_lock();
gx_powerdown(info);
fb_set_suspend(info, 1);
- release_console_sem();
+ console_unlock();
}
/* there's no point in setting PCI states; we emulate PCI, so
@@ -361,7 +361,7 @@ static int gxfb_resume(struct pci_dev *pdev)
struct fb_info *info = pci_get_drvdata(pdev);
int ret;
- acquire_console_sem();
+ console_lock();
ret = gx_powerup(info);
if (ret) {
printk(KERN_ERR "gxfb: power up failed!\n");
@@ -369,7 +369,7 @@ static int gxfb_resume(struct pci_dev *pdev)
}
fb_set_suspend(info, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
#endif
diff --git a/drivers/video/geode/lxfb_core.c b/drivers/video/geode/lxfb_core.c
index 39bdbed..416851c 100644
--- a/drivers/video/geode/lxfb_core.c
+++ b/drivers/video/geode/lxfb_core.c
@@ -465,10 +465,10 @@ static int lxfb_suspend(struct pci_dev *pdev, pm_message_t state)
struct fb_info *info = pci_get_drvdata(pdev);
if (state.event == PM_EVENT_SUSPEND) {
- acquire_console_sem();
+ console_lock();
lx_powerdown(info);
fb_set_suspend(info, 1);
- release_console_sem();
+ console_unlock();
}
/* there's no point in setting PCI states; we emulate PCI, so
@@ -482,7 +482,7 @@ static int lxfb_resume(struct pci_dev *pdev)
struct fb_info *info = pci_get_drvdata(pdev);
int ret;
- acquire_console_sem();
+ console_lock();
ret = lx_powerup(info);
if (ret) {
printk(KERN_ERR "lxfb: power up failed!\n");
@@ -490,7 +490,7 @@ static int lxfb_resume(struct pci_dev *pdev)
}
fb_set_suspend(info, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
#else
diff --git a/drivers/video/i810/i810_main.c b/drivers/video/i810/i810_main.c
index 5743ea2..318f6fb 100644
--- a/drivers/video/i810/i810_main.c
+++ b/drivers/video/i810/i810_main.c
@@ -1574,7 +1574,7 @@ static int i810fb_suspend(struct pci_dev *dev, pm_message_t mesg)
return 0;
}
- acquire_console_sem();
+ console_lock();
fb_set_suspend(info, 1);
if (info->fbops->fb_sync)
@@ -1587,7 +1587,7 @@ static int i810fb_suspend(struct pci_dev *dev, pm_message_t mesg)
pci_save_state(dev);
pci_disable_device(dev);
pci_set_power_state(dev, pci_choose_state(dev, mesg));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1605,7 +1605,7 @@ static int i810fb_resume(struct pci_dev *dev)
return 0;
}
- acquire_console_sem();
+ console_lock();
pci_set_power_state(dev, PCI_D0);
pci_restore_state(dev);
@@ -1621,7 +1621,7 @@ static int i810fb_resume(struct pci_dev *dev)
fb_set_suspend (info, 0);
info->fbops->fb_blank(VESA_NO_BLANKING, info);
fail:
- release_console_sem();
+ console_unlock();
return 0;
}
/***********************************************************************
diff --git a/drivers/video/jz4740_fb.c b/drivers/video/jz4740_fb.c
index 670ecaa..de36693 100644
--- a/drivers/video/jz4740_fb.c
+++ b/drivers/video/jz4740_fb.c
@@ -778,9 +778,9 @@ static int jzfb_suspend(struct device *dev)
{
struct jzfb *jzfb = dev_get_drvdata(dev);
- acquire_console_sem();
+ console_lock();
fb_set_suspend(jzfb->fb, 1);
- release_console_sem();
+ console_unlock();
mutex_lock(&jzfb->lock);
if (jzfb->is_enabled)
@@ -800,9 +800,9 @@ static int jzfb_resume(struct device *dev)
jzfb_enable(jzfb);
mutex_unlock(&jzfb->lock);
- acquire_console_sem();
+ console_lock();
fb_set_suspend(jzfb->fb, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/mx3fb.c b/drivers/video/mx3fb.c
index cb01391..7e3a490 100644
--- a/drivers/video/mx3fb.c
+++ b/drivers/video/mx3fb.c
@@ -1177,9 +1177,9 @@ static int mx3fb_suspend(struct platform_device *pdev, pm_message_t state)
struct mx3fb_data *mx3fb = platform_get_drvdata(pdev);
struct mx3fb_info *mx3_fbi = mx3fb->fbi->par;
- acquire_console_sem();
+ console_lock();
fb_set_suspend(mx3fb->fbi, 1);
- release_console_sem();
+ console_unlock();
if (mx3_fbi->blank == FB_BLANK_UNBLANK) {
sdc_disable_channel(mx3_fbi);
@@ -1202,9 +1202,9 @@ static int mx3fb_resume(struct platform_device *pdev)
sdc_set_brightness(mx3fb, mx3fb->backlight_level);
}
- acquire_console_sem();
+ console_lock();
fb_set_suspend(mx3fb->fbi, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/nvidia/nvidia.c b/drivers/video/nvidia/nvidia.c
index efe10ff..081dc47 100644
--- a/drivers/video/nvidia/nvidia.c
+++ b/drivers/video/nvidia/nvidia.c
@@ -1057,7 +1057,7 @@ static int nvidiafb_suspend(struct pci_dev *dev, pm_message_t mesg)
if (mesg.event == PM_EVENT_PRETHAW)
mesg.event = PM_EVENT_FREEZE;
- acquire_console_sem();
+ console_lock();
par->pm_state = mesg.event;
if (mesg.event & PM_EVENT_SLEEP) {
@@ -1070,7 +1070,7 @@ static int nvidiafb_suspend(struct pci_dev *dev, pm_message_t mesg)
}
dev->dev.power.power_state = mesg;
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1079,7 +1079,7 @@ static int nvidiafb_resume(struct pci_dev *dev)
struct fb_info *info = pci_get_drvdata(dev);
struct nvidia_par *par = info->par;
- acquire_console_sem();
+ console_lock();
pci_set_power_state(dev, PCI_D0);
if (par->pm_state != PM_EVENT_FREEZE) {
@@ -1097,7 +1097,7 @@ static int nvidiafb_resume(struct pci_dev *dev)
nvidiafb_blank(FB_BLANK_UNBLANK, info);
fail:
- release_console_sem();
+ console_unlock();
return 0;
}
#else
diff --git a/drivers/video/ps3fb.c b/drivers/video/ps3fb.c
index 9c0144e..65560a1 100644
--- a/drivers/video/ps3fb.c
+++ b/drivers/video/ps3fb.c
@@ -513,9 +513,9 @@ static int ps3fb_release(struct fb_info *info, int user)
if (atomic_dec_and_test(&ps3fb.f_count)) {
if (atomic_read(&ps3fb.ext_flip)) {
atomic_set(&ps3fb.ext_flip, 0);
- if (!try_acquire_console_sem()) {
+ if (console_trylock()) {
ps3fb_sync(info, 0); /* single buffer */
- release_console_sem();
+ console_unlock();
}
}
}
@@ -830,14 +830,14 @@ static int ps3fb_ioctl(struct fb_info *info, unsigned int cmd,
if (vmode) {
var = info->var;
fb_videomode_to_var(&var, vmode);
- acquire_console_sem();
+ console_lock();
info->flags |= FBINFO_MISC_USEREVENT;
/* Force, in case only special bits changed */
var.activate |= FB_ACTIVATE_FORCE;
par->new_mode_id = val;
retval = fb_set_var(info, &var);
info->flags &= ~FBINFO_MISC_USEREVENT;
- release_console_sem();
+ console_unlock();
}
break;
}
@@ -881,9 +881,9 @@ static int ps3fb_ioctl(struct fb_info *info, unsigned int cmd,
break;
dev_dbg(info->device, "PS3FB_IOCTL_FSEL:%d\n", val);
- acquire_console_sem();
+ console_lock();
retval = ps3fb_sync(info, val);
- release_console_sem();
+ console_unlock();
break;
default:
@@ -903,9 +903,9 @@ static int ps3fbd(void *arg)
set_current_state(TASK_INTERRUPTIBLE);
if (ps3fb.is_kicked) {
ps3fb.is_kicked = 0;
- acquire_console_sem();
+ console_lock();
ps3fb_sync(info, 0); /* single buffer */
- release_console_sem();
+ console_unlock();
}
schedule();
}
diff --git a/drivers/video/s3fb.c b/drivers/video/s3fb.c
index dce8c97..75738a9 100644
--- a/drivers/video/s3fb.c
+++ b/drivers/video/s3fb.c
@@ -22,7 +22,7 @@
#include <linux/svga.h>
#include <linux/init.h>
#include <linux/pci.h>
-#include <linux/console.h> /* Why should fb driver call console functions? because acquire_console_sem() */
+#include <linux/console.h> /* Why should fb driver call console functions? because console_lock() */
#include <video/vga.h>
#ifdef CONFIG_MTRR
@@ -1113,12 +1113,12 @@ static int s3_pci_suspend(struct pci_dev* dev, pm_message_t state)
dev_info(info->device, "suspend\n");
- acquire_console_sem();
+ console_lock();
mutex_lock(&(par->open_lock));
if ((state.event == PM_EVENT_FREEZE) || (par->ref_count == 0)) {
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1129,7 +1129,7 @@ static int s3_pci_suspend(struct pci_dev* dev, pm_message_t state)
pci_set_power_state(dev, pci_choose_state(dev, state));
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1145,12 +1145,12 @@ static int s3_pci_resume(struct pci_dev* dev)
dev_info(info->device, "resume\n");
- acquire_console_sem();
+ console_lock();
mutex_lock(&(par->open_lock));
if (par->ref_count == 0) {
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -1159,7 +1159,7 @@ static int s3_pci_resume(struct pci_dev* dev)
err = pci_enable_device(dev);
if (err) {
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
dev_err(info->device, "error %d enabling device for resume\n", err);
return err;
}
@@ -1169,7 +1169,7 @@ static int s3_pci_resume(struct pci_dev* dev)
fb_set_suspend(info, 0);
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/savage/savagefb_driver.c b/drivers/video/savage/savagefb_driver.c
index 842d157..487911e 100644
--- a/drivers/video/savage/savagefb_driver.c
+++ b/drivers/video/savage/savagefb_driver.c
@@ -2373,7 +2373,7 @@ static int savagefb_suspend(struct pci_dev *dev, pm_message_t mesg)
if (mesg.event == PM_EVENT_FREEZE)
return 0;
- acquire_console_sem();
+ console_lock();
fb_set_suspend(info, 1);
if (info->fbops->fb_sync)
@@ -2385,7 +2385,7 @@ static int savagefb_suspend(struct pci_dev *dev, pm_message_t mesg)
pci_save_state(dev);
pci_disable_device(dev);
pci_set_power_state(dev, pci_choose_state(dev, mesg));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -2409,7 +2409,7 @@ static int savagefb_resume(struct pci_dev* dev)
return 0;
}
- acquire_console_sem();
+ console_lock();
pci_set_power_state(dev, PCI_D0);
pci_restore_state(dev);
@@ -2423,7 +2423,7 @@ static int savagefb_resume(struct pci_dev* dev)
savagefb_set_par(info);
fb_set_suspend(info, 0);
savagefb_blank(FB_BLANK_UNBLANK, info);
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/sh_mobile_hdmi.c b/drivers/video/sh_mobile_hdmi.c
index 74d9f54..2b9e56a 100644
--- a/drivers/video/sh_mobile_hdmi.c
+++ b/drivers/video/sh_mobile_hdmi.c
@@ -1151,7 +1151,7 @@ static void sh_hdmi_edid_work_fn(struct work_struct *work)
ch = info->par;
- acquire_console_sem();
+ console_lock();
/* HDMI plug in */
if (!sh_hdmi_must_reconfigure(hdmi) &&
@@ -1171,7 +1171,7 @@ static void sh_hdmi_edid_work_fn(struct work_struct *work)
fb_set_suspend(info, 0);
}
- release_console_sem();
+ console_unlock();
} else {
ret = 0;
if (!hdmi->info)
@@ -1181,12 +1181,12 @@ static void sh_hdmi_edid_work_fn(struct work_struct *work)
fb_destroy_modedb(hdmi->monspec.modedb);
hdmi->monspec.modedb = NULL;
- acquire_console_sem();
+ console_lock();
/* HDMI disconnect */
fb_set_suspend(hdmi->info, 1);
- release_console_sem();
+ console_unlock();
pm_runtime_put(hdmi->dev);
}
diff --git a/drivers/video/sh_mobile_lcdcfb.c b/drivers/video/sh_mobile_lcdcfb.c
index bd4840a..bf12e53 100644
--- a/drivers/video/sh_mobile_lcdcfb.c
+++ b/drivers/video/sh_mobile_lcdcfb.c
@@ -912,9 +912,9 @@ static int sh_mobile_release(struct fb_info *info, int user)
/* Nothing to reconfigure, when called from fbcon */
if (user) {
- acquire_console_sem();
+ console_lock();
sh_mobile_fb_reconfig(info);
- release_console_sem();
+ console_unlock();
}
mutex_unlock(&ch->open_lock);
diff --git a/drivers/video/sm501fb.c b/drivers/video/sm501fb.c
index b7dc180..bcb44a5 100644
--- a/drivers/video/sm501fb.c
+++ b/drivers/video/sm501fb.c
@@ -2010,9 +2010,9 @@ static int sm501fb_suspend_fb(struct sm501fb_info *info,
/* tell console/fb driver we are suspending */
- acquire_console_sem();
+ console_lock();
fb_set_suspend(fbi, 1);
- release_console_sem();
+ console_unlock();
/* backup copies in case chip is powered down over suspend */
@@ -2069,9 +2069,9 @@ static void sm501fb_resume_fb(struct sm501fb_info *info,
memcpy_toio(par->cursor.k_addr, par->store_cursor,
par->cursor.size);
- acquire_console_sem();
+ console_lock();
fb_set_suspend(fbi, 0);
- release_console_sem();
+ console_unlock();
vfree(par->store_fb);
vfree(par->store_cursor);
diff --git a/drivers/video/tmiofb.c b/drivers/video/tmiofb.c
index 6913fe1..dfef88c 100644
--- a/drivers/video/tmiofb.c
+++ b/drivers/video/tmiofb.c
@@ -25,7 +25,7 @@
#include <linux/fb.h>
#include <linux/interrupt.h>
#include <linux/delay.h>
-/* Why should fb driver call console functions? because acquire_console_sem() */
+/* Why should fb driver call console functions? because console_lock() */
#include <linux/console.h>
#include <linux/mfd/core.h>
#include <linux/mfd/tmio.h>
@@ -944,7 +944,7 @@ static int tmiofb_suspend(struct platform_device *dev, pm_message_t state)
struct mfd_cell *cell = dev->dev.platform_data;
int retval = 0;
- acquire_console_sem();
+ console_lock();
fb_set_suspend(info, 1);
@@ -965,7 +965,7 @@ static int tmiofb_suspend(struct platform_device *dev, pm_message_t state)
if (cell->suspend)
retval = cell->suspend(dev);
- release_console_sem();
+ console_unlock();
return retval;
}
@@ -976,7 +976,7 @@ static int tmiofb_resume(struct platform_device *dev)
struct mfd_cell *cell = dev->dev.platform_data;
int retval = 0;
- acquire_console_sem();
+ console_lock();
if (cell->resume) {
retval = cell->resume(dev);
@@ -992,7 +992,7 @@ static int tmiofb_resume(struct platform_device *dev)
fb_set_suspend(info, 0);
out:
- release_console_sem();
+ console_unlock();
return retval;
}
#else
diff --git a/drivers/video/via/viafbdev.c b/drivers/video/via/viafbdev.c
index 289edd5..4e66349 100644
--- a/drivers/video/via/viafbdev.c
+++ b/drivers/video/via/viafbdev.c
@@ -1674,17 +1674,17 @@ static int parse_mode(const char *str, u32 *xres, u32 *yres)
#ifdef CONFIG_PM
static int viafb_suspend(void *unused)
{
- acquire_console_sem();
+ console_lock();
fb_set_suspend(viafbinfo, 1);
viafb_sync(viafbinfo);
- release_console_sem();
+ console_unlock();
return 0;
}
static int viafb_resume(void *unused)
{
- acquire_console_sem();
+ console_lock();
if (viaparinfo->shared->vdev->engine_mmio)
viafb_reset_engine(viaparinfo);
viafb_set_par(viafbinfo);
@@ -1692,7 +1692,7 @@ static int viafb_resume(void *unused)
viafb_set_par(viafbinfo1);
fb_set_suspend(viafbinfo, 0);
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/vt8623fb.c b/drivers/video/vt8623fb.c
index 85d76ec..a2965ab 100644
--- a/drivers/video/vt8623fb.c
+++ b/drivers/video/vt8623fb.c
@@ -23,7 +23,7 @@
#include <linux/svga.h>
#include <linux/init.h>
#include <linux/pci.h>
-#include <linux/console.h> /* Why should fb driver call console functions? because acquire_console_sem() */
+#include <linux/console.h> /* Why should fb driver call console functions? because console_lock() */
#include <video/vga.h>
#ifdef CONFIG_MTRR
@@ -819,12 +819,12 @@ static int vt8623_pci_suspend(struct pci_dev* dev, pm_message_t state)
dev_info(info->device, "suspend\n");
- acquire_console_sem();
+ console_lock();
mutex_lock(&(par->open_lock));
if ((state.event == PM_EVENT_FREEZE) || (par->ref_count == 0)) {
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -835,7 +835,7 @@ static int vt8623_pci_suspend(struct pci_dev* dev, pm_message_t state)
pci_set_power_state(dev, pci_choose_state(dev, state));
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
@@ -850,7 +850,7 @@ static int vt8623_pci_resume(struct pci_dev* dev)
dev_info(info->device, "resume\n");
- acquire_console_sem();
+ console_lock();
mutex_lock(&(par->open_lock));
if (par->ref_count == 0)
@@ -869,7 +869,7 @@ static int vt8623_pci_resume(struct pci_dev* dev)
fail:
mutex_unlock(&(par->open_lock));
- release_console_sem();
+ console_unlock();
return 0;
}
diff --git a/drivers/video/xen-fbfront.c b/drivers/video/xen-fbfront.c
index 3e6934d..a20218c 100644
--- a/drivers/video/xen-fbfront.c
+++ b/drivers/video/xen-fbfront.c
@@ -491,12 +491,12 @@ xenfb_make_preferred_console(void)
if (console_set_on_cmdline)
return;
- acquire_console_sem();
+ console_lock();
for_each_console(c) {
if (!strcmp(c->name, "tty") && c->index == 0)
break;
}
- release_console_sem();
+ console_unlock();
if (c) {
unregister_console(c);
c->flags |= CON_CONSDEV;
diff --git a/fs/proc/consoles.c b/fs/proc/consoles.c
index eafc22a..b701eaa 100644
--- a/fs/proc/consoles.c
+++ b/fs/proc/consoles.c
@@ -67,7 +67,7 @@ static void *c_start(struct seq_file *m, loff_t *pos)
struct console *con;
loff_t off = 0;
- acquire_console_sem();
+ console_lock();
for_each_console(con)
if (off++ == *pos)
break;
@@ -84,7 +84,7 @@ static void *c_next(struct seq_file *m, void *v, loff_t *pos)
static void c_stop(struct seq_file *m, void *v)
{
- release_console_sem();
+ console_unlock();
}
static const struct seq_operations consoles_op = {
diff --git a/include/linux/console.h b/include/linux/console.h
index 9774fe6..7453cfd 100644
--- a/include/linux/console.h
+++ b/include/linux/console.h
@@ -139,9 +139,9 @@ extern int update_console_cmdline(char *name, int idx, char *name_new, int idx_n
extern void register_console(struct console *);
extern int unregister_console(struct console *);
extern struct console *console_drivers;
-extern void acquire_console_sem(void);
-extern int try_acquire_console_sem(void);
-extern void release_console_sem(void);
+extern void console_lock(void);
+extern int console_trylock(void);
+extern void console_unlock(void);
extern void console_conditional_schedule(void);
extern void console_unblank(void);
extern struct tty_driver *console_device(int *);
diff --git a/kernel/printk.c b/kernel/printk.c
index 53d9a9e..ef85a8e 100644
--- a/kernel/printk.c
+++ b/kernel/printk.c
@@ -97,7 +97,7 @@ static int console_locked, console_suspended;
/*
* logbuf_lock protects log_buf, log_start, log_end, con_start and logged_chars
* It is also used in interesting ways to provide interlocking in
- * release_console_sem().
+ * console_unlock();.
*/
static DEFINE_SPINLOCK(logbuf_lock);
@@ -501,7 +501,7 @@ static void _call_console_drivers(unsigned start,
/*
* Call the console drivers, asking them to write out
* log_buf[start] to log_buf[end - 1].
- * The console_sem must be held.
+ * The console_lock must be held.
*/
static void call_console_drivers(unsigned start, unsigned end)
{
@@ -604,11 +604,11 @@ static int have_callable_console(void)
*
* This is printk(). It can be called from any context. We want it to work.
*
- * We try to grab the console_sem. If we succeed, it's easy - we log the output and
+ * We try to grab the console_lock. If we succeed, it's easy - we log the output and
* call the console drivers. If we fail to get the semaphore we place the output
* into the log buffer and return. The current holder of the console_sem will
- * notice the new output in release_console_sem() and will send it to the
- * consoles before releasing the semaphore.
+ * notice the new output in console_unlock(); and will send it to the
+ * consoles before releasing the lock.
*
* One effect of this deferred printing is that code which calls printk() and
* then changes console_loglevel may break. This is because console_loglevel
@@ -659,19 +659,19 @@ static inline int can_use_console(unsigned int cpu)
/*
* Try to get console ownership to actually show the kernel
* messages from a 'printk'. Return true (and with the
- * console_semaphore held, and 'console_locked' set) if it
+ * console_lock held, and 'console_locked' set) if it
* is successful, false otherwise.
*
* This gets called with the 'logbuf_lock' spinlock held and
* interrupts disabled. It should return with 'lockbuf_lock'
* released but interrupts still disabled.
*/
-static int acquire_console_semaphore_for_printk(unsigned int cpu)
+static int console_trylock_for_printk(unsigned int cpu)
__releases(&logbuf_lock)
{
int retval = 0;
- if (!try_acquire_console_sem()) {
+ if (console_trylock()) {
retval = 1;
/*
@@ -827,12 +827,12 @@ asmlinkage int vprintk(const char *fmt, va_list args)
* actual magic (print out buffers, wake up klogd,
* etc).
*
- * The acquire_console_semaphore_for_printk() function
+ * The console_trylock_for_printk() function
* will release 'logbuf_lock' regardless of whether it
* actually gets the semaphore or not.
*/
- if (acquire_console_semaphore_for_printk(this_cpu))
- release_console_sem();
+ if (console_trylock_for_printk(this_cpu))
+ console_unlock();
lockdep_on();
out_restore_irqs:
@@ -993,7 +993,7 @@ void suspend_console(void)
if (!console_suspend_enabled)
return;
printk("Suspending console(s) (use no_console_suspend to debug)\n");
- acquire_console_sem();
+ console_lock();;
console_suspended = 1;
up(&console_sem);
}
@@ -1004,7 +1004,7 @@ void resume_console(void)
return;
down(&console_sem);
console_suspended = 0;
- release_console_sem();
+ console_unlock();
}
/**
@@ -1027,21 +1027,21 @@ static int __cpuinit console_cpu_notify(struct notifier_block *self,
case CPU_DYING:
case CPU_DOWN_FAILED:
case CPU_UP_CANCELED:
- acquire_console_sem();
- release_console_sem();
+ console_lock();;
+ console_unlock();
}
return NOTIFY_OK;
}
/**
- * acquire_console_sem - lock the console system for exclusive use.
+ * console_lock - lock the console system for exclusive use.
*
- * Acquires a semaphore which guarantees that the caller has
+ * Acquires a lock which guarantees that the caller has
* exclusive access to the console system and the console_drivers list.
*
* Can sleep, returns nothing.
*/
-void acquire_console_sem(void)
+void console_lock(void)
{
BUG_ON(in_interrupt());
down(&console_sem);
@@ -1050,21 +1050,29 @@ void acquire_console_sem(void)
console_locked = 1;
console_may_schedule = 1;
}
-EXPORT_SYMBOL(acquire_console_sem);
+EXPORT_SYMBOL(console_lock);
-int try_acquire_console_sem(void)
+/**
+ * console_trylock - try to lock the console system for exclusive use.
+ *
+ * Tried to acquire a lock which guarantees that the caller has
+ * exclusive access to the console system and the console_drivers list.
+ *
+ * returns -1 on success, and 0 on failure to acquire the lock.
+ */
+int console_trylock(void)
{
if (down_trylock(&console_sem))
- return -1;
+ return 0;
if (console_suspended) {
up(&console_sem);
- return -1;
+ return 0;
}
console_locked = 1;
console_may_schedule = 0;
- return 0;
+ return -1;
}
-EXPORT_SYMBOL(try_acquire_console_sem);
+EXPORT_SYMBOL(console_trylock);
int is_console_locked(void)
{
@@ -1095,20 +1103,20 @@ void wake_up_klogd(void)
}
/**
- * release_console_sem - unlock the console system
+ * console_unlock - unlock the console system
*
- * Releases the semaphore which the caller holds on the console system
+ * Releases the console_lock which the caller holds on the console system
* and the console driver list.
*
- * While the semaphore was held, console output may have been buffered
- * by printk(). If this is the case, release_console_sem() emits
- * the output prior to releasing the semaphore.
+ * While the console_lock was held, console output may have been buffered
+ * by printk(). If this is the case, console_unlock(); emits
+ * the output prior to releasing the lock.
*
* If there is output waiting for klogd, we wake it up.
*
- * release_console_sem() may be called from any context.
+ * console_unlock(); may be called from any context.
*/
-void release_console_sem(void)
+void console_unlock(void)
{
unsigned long flags;
unsigned _con_start, _log_end;
@@ -1141,7 +1149,7 @@ void release_console_sem(void)
if (wake_klogd)
wake_up_klogd();
}
-EXPORT_SYMBOL(release_console_sem);
+EXPORT_SYMBOL(console_unlock);
/**
* console_conditional_schedule - yield the CPU if required
@@ -1150,7 +1158,7 @@ EXPORT_SYMBOL(release_console_sem);
* if this CPU should yield the CPU to another task, do
* so here.
*
- * Must be called within acquire_console_sem().
+ * Must be called within console_lock();.
*/
void __sched console_conditional_schedule(void)
{
@@ -1171,14 +1179,14 @@ void console_unblank(void)
if (down_trylock(&console_sem) != 0)
return;
} else
- acquire_console_sem();
+ console_lock();
console_locked = 1;
console_may_schedule = 0;
for_each_console(c)
if ((c->flags & CON_ENABLED) && c->unblank)
c->unblank();
- release_console_sem();
+ console_unlock();
}
/*
@@ -1189,7 +1197,7 @@ struct tty_driver *console_device(int *index)
struct console *c;
struct tty_driver *driver = NULL;
- acquire_console_sem();
+ console_lock();
for_each_console(c) {
if (!c->device)
continue;
@@ -1197,7 +1205,7 @@ struct tty_driver *console_device(int *index)
if (driver)
break;
}
- release_console_sem();
+ console_unlock();
return driver;
}
@@ -1208,17 +1216,17 @@ struct tty_driver *console_device(int *index)
*/
void console_stop(struct console *console)
{
- acquire_console_sem();
+ console_lock();
console->flags &= ~CON_ENABLED;
- release_console_sem();
+ console_unlock();
}
EXPORT_SYMBOL(console_stop);
void console_start(struct console *console)
{
- acquire_console_sem();
+ console_lock();
console->flags |= CON_ENABLED;
- release_console_sem();
+ console_unlock();
}
EXPORT_SYMBOL(console_start);
@@ -1340,7 +1348,7 @@ void register_console(struct console *newcon)
* Put this console in the list - keep the
* preferred driver at the head of the list.
*/
- acquire_console_sem();
+ console_lock();
if ((newcon->flags & CON_CONSDEV) || console_drivers == NULL) {
newcon->next = console_drivers;
console_drivers = newcon;
@@ -1352,14 +1360,14 @@ void register_console(struct console *newcon)
}
if (newcon->flags & CON_PRINTBUFFER) {
/*
- * release_console_sem() will print out the buffered messages
+ * console_unlock(); will print out the buffered messages
* for us.
*/
spin_lock_irqsave(&logbuf_lock, flags);
con_start = log_start;
spin_unlock_irqrestore(&logbuf_lock, flags);
}
- release_console_sem();
+ console_unlock();
console_sysfs_notify();
/*
@@ -1396,7 +1404,7 @@ int unregister_console(struct console *console)
return braille_unregister_console(console);
#endif
- acquire_console_sem();
+ console_lock();
if (console_drivers == console) {
console_drivers=console->next;
res = 0;
@@ -1418,7 +1426,7 @@ int unregister_console(struct console *console)
if (console_drivers != NULL && console->flags & CON_CONSDEV)
console_drivers->flags |= CON_CONSDEV;
- release_console_sem();
+ console_unlock();
console_sysfs_notify();
return res;
}
--
1.7.2.3
^ permalink raw reply related
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Srivatsa Vaddagiri @ 2011-01-20 14:34 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <1295530906.28776.171.camel@laptop>
On Thu, Jan 20, 2011 at 02:41:46PM +0100, Peter Zijlstra wrote:
> On Thu, 2011-01-20 at 17:29 +0530, Srivatsa Vaddagiri wrote:
> >
> > If we had a yield-to [1] sort of interface _and_ information on which vcpu
> > owns a lock, then lock-spinners can yield-to the owning vcpu,
>
> and then I'd nak it for being stupid ;-)
>
> really, yield*() is retarded, never even consider using it. If you've
> got the actual owner you can do full blown PI, which is tons better than
> a 'do-something-random' call.
Yes definitely that would be much better than yield-to.
> The only reason the whole non-virt pause loop filtering muck uses it is
> because it really doesn't know anything, and do-something is pretty much
> all it can do. Its a broken interface.
- vatsa
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Peter Zijlstra @ 2011-01-20 13:41 UTC (permalink / raw)
To: vatsa
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110120115958.GB11177@linux.vnet.ibm.com>
On Thu, 2011-01-20 at 17:29 +0530, Srivatsa Vaddagiri wrote:
>
> If we had a yield-to [1] sort of interface _and_ information on which vcpu
> owns a lock, then lock-spinners can yield-to the owning vcpu,
and then I'd nak it for being stupid ;-)
really, yield*() is retarded, never even consider using it. If you've
got the actual owner you can do full blown PI, which is tons better than
a 'do-something-random' call.
The only reason the whole non-virt pause loop filtering muck uses it is
because it really doesn't know anything, and do-something is pretty much
all it can do. Its a broken interface.
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Srivatsa Vaddagiri @ 2011-01-20 11:59 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <4D373340.60608@goop.org>
On Wed, Jan 19, 2011 at 10:53:52AM -0800, Jeremy Fitzhardinge wrote:
> > I didn't really read the patch, and I totally forgot everything from
> > when I looked at the Xen series, but does the Xen/KVM hypercall
> > interface for this include the vcpu to await the kick from?
> >
> > My guess is not, since the ticket locks used don't know who the owner
> > is, which is of course, sad. There are FIFO spinlock implementations
> > that can do this though.. although I think they all have a bigger memory
> > footprint.
>
> At least in the Xen code, a current owner isn't very useful, because we
> need the current owner to kick the *next* owner to life at release time,
> which we can't do without some structure recording which ticket belongs
> to which cpu.
If we had a yield-to [1] sort of interface _and_ information on which vcpu
owns a lock, then lock-spinners can yield-to the owning vcpu, while the
unlocking vcpu can yield-to the next-vcpu-in-waiting. The key here is not to
sleep when waiting for locks (as implemented by current patch-series, which can
put other VMs at an advantage by giving them more time than they are entitled
to) and also to ensure that lock-owner as well as the next-in-line lock-owner
are not unduly made to wait for cpu.
Is there a way we can dynamically expand the size of lock only upon contention
to include additional information like owning vcpu? Have the lock point to a
per-cpu area upon contention where additional details can be stored perhaps?
1. https://lkml.org/lkml/2011/1/14/44
- vatsa
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Srivatsa Vaddagiri @ 2011-01-20 11:42 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <4D373340.60608@goop.org>
On Wed, Jan 19, 2011 at 10:53:52AM -0800, Jeremy Fitzhardinge wrote:
> > The reason for wanting this should be clear I guess, it allows PI.
>
> Well, if we can expand the spinlock to include an owner, then all this
> becomes moot...
How so? Having an owner will not eliminate the need for pv-ticketlocks
afaict. We still need a mechanism to reduce latency in scheduling the next
thread-in-waiting, which is achieved by your patches.
- vatsa
^ permalink raw reply
* [PULL] lguest and virtio.
From: Rusty Russell @ 2011-01-20 11:13 UTC (permalink / raw)
To: Linus Torvalds; +Cc: lguest, linux-kernel, virtualization
The following changes since commit 8a335bc631ac9c43675821580c26ebf95a3044ba:
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/nab/scsi-post-merge-2.6 (2011-01-16 15:06:43 -0800)
are available in the git repository at:
ssh://master.kernel.org/pub/scm/linux/kernel/git/rusty/linux-2.6-for-linus.git master
Christoph Lameter (1):
lguest: Use this_cpu_ops
Milton Miller (1):
virtio: remove virtio-pci root device
Philip Sanderson (3):
lguest: --username and --chroot options
lguest: example launcher to use guard pages, drop PROT_EXEC, fix limit logic
lguest: document --rng in example Launcher
Randy Dunlap (1):
LGUEST_GUEST: fix unmet direct dependencies (VIRTUALIZATION && VIRTIO)
Rusty Russell (1):
lguest: compile fixes
Documentation/lguest/lguest.c | 73 ++++++++++++++++++++++++++++++++++----
Documentation/lguest/lguest.txt | 5 +++
arch/x86/lguest/Kconfig | 1 +
arch/x86/lguest/boot.c | 2 +-
drivers/lguest/page_tables.c | 2 +-
drivers/lguest/x86/core.c | 4 +-
drivers/virtio/virtio_pci.c | 20 +---------
7 files changed, 77 insertions(+), 30 deletions(-)
^ permalink raw reply
* Re: [PATCH 13/14] x86/ticketlock: add slowpath logic
From: Jan Beulich @ 2011-01-20 9:52 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Jeremy Fitzhardinge, Mathieu Desnoyers, Nick Piggin,
Peter Zijlstra, Linux Kernel Mailing List, vatsa, Eric Dumazet,
Xen-devel, suzuki, Avi Kivity, H. Peter Anvin, xiyou.wangcong,
Linux Virtualization
In-Reply-To: <4D37338E.8030509@goop.org>
>>> On 19.01.11 at 19:55, Jeremy Fitzhardinge <jeremy@goop.org> wrote:
> On 01/19/2011 10:39 AM, Srivatsa Vaddagiri wrote:
>> I have tested quite extensively with booting a 16-vcpu guest (on a 16-pcpu
> host)
>> and running kernel compine (with 32-threads). Without this patch, I had
>> difficulty booting/shutting-down successfully (it would hang mid-way).
>
> Sounds good. But I like to test with "make -j 100-200" to really give
> things a workout.
Hmm, in my experience, heavily over-committing CPUs (e.g. a
domain with double or more the vCPU-s that the system has
pCPU-s, or the pCPU-s the domain is allowed to run on) is a
much better test for eventual problems in the spin lock code
paths.
Jan
^ permalink raw reply
* Re: Flow Control and Port Mirroring Revisited
From: Simon Horman @ 2011-01-20 8:38 UTC (permalink / raw)
To: Rick Jones
Cc: Michael S. Tsirkin, Jesse Gross, Rusty Russell, virtualization,
dev, virtualization, netdev, kvm
In-Reply-To: <4D35ECE2.4040901@hp.com>
[ Trimmed Eric from CC list as vger was complaining that it is too long ]
On Tue, Jan 18, 2011 at 11:41:22AM -0800, Rick Jones wrote:
> >So it won't be all that simple to implement well, and before we try,
> >I'd like to know whether there are applications that are helped
> >by it. For example, we could try to measure latency at various
> >pps and see whether the backpressure helps. netperf has -b, -w
> >flags which might help these measurements.
>
> Those options are enabled when one adds --enable-burst to the
> pre-compilation ./configure of netperf (one doesn't have to
> recompile netserver). However, if one is also looking at latency
> statistics via the -j option in the top-of-trunk, or simply at the
> histogram with --enable-histogram on the ./configure and a verbosity
> level of 2 (global -v 2) then one wants the very top of trunk
> netperf from:
Hi,
I have constructed a test where I run an un-paced UDP_STREAM test in
one guest and a paced omni rr test in another guest at the same time.
Breifly I get the following results from the omni test..
1. Omni test only: MEAN_LATENCY=272.00
2. Omni and stream test: MEAN_LATENCY=3423.00
3. cpu and net_cls group: MEAN_LATENCY=493.00
As per 2 plus cgoups are created for each guest
and guest tasks added to the groups
4. 100Mbit/s class: MEAN_LATENCY=273.00
As per 3 plus the net_cls groups each have a 100MBit/s HTB class
5. cpu.shares=128: MEAN_LATENCY=652.00
As per 4 plus the cpu groups have cpu.shares set to 128
6. Busy CPUS: MEAN_LATENCY=15126.00
As per 5 but the CPUs are made busy using a simple shell while loop
There is a bit of noise in the results as the two netperf invocations
aren't started at exactly the same moment
For reference, my netperf invocations are:
netperf -c -C -t UDP_STREAM -H 172.17.60.216 -l 12
netperf.omni -p 12866 -D -c -C -H 172.17.60.216 -t omni -j -v 2 -- -r 1 -d rr -k foo -b 1 -w 200 -m 200
foo contains
PROTOCOL
THROUGHPUT,THROUGHPUT_UNITS
LOCAL_SEND_THROUGHPUT
LOCAL_RECV_THROUGHPUT
REMOTE_SEND_THROUGHPUT
REMOTE_RECV_THROUGHPUT
RT_LATENCY,MIN_LATENCY,MEAN_LATENCY,MAX_LATENCY
P50_LATENCY,P90_LATENCY,P99_LATENCY,STDDEV_LATENCY
LOCAL_CPU_UTIL,REMOTE_CPU_UTIL
^ permalink raw reply
* Re: [PATCH 13/14] x86/ticketlock: add slowpath logic
From: Srivatsa Vaddagiri @ 2011-01-20 4:28 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <4D37338E.8030509@goop.org>
On Wed, Jan 19, 2011 at 10:55:10AM -0800, Jeremy Fitzhardinge wrote:
> On 01/19/2011 10:39 AM, Srivatsa Vaddagiri wrote:
> > I have tested quite extensively with booting a 16-vcpu guest (on a 16-pcpu host)
> > and running kernel compine (with 32-threads). Without this patch, I had
> > difficulty booting/shutting-down successfully (it would hang mid-way).
>
> Sounds good. But I like to test with "make -j 100-200" to really give
> things a workout.
I reran the tests with the following configuration:
Host : 16pcpus, 32GB Memory
Single KVM guest w/ 32vcpus, 16GB memory
It survived a kernel compile with 200 threads once and another time with 300
threads. 'make allyesconfig' was run before compiling the kernel. So looks
quite stable to me ..additional testing from wider community will also help.
- vasa
^ permalink raw reply
* Re: [PATCH 13/14] x86/ticketlock: add slowpath logic
From: Jeremy Fitzhardinge @ 2011-01-19 18:55 UTC (permalink / raw)
To: vatsa
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119183914.GB7235@linux.vnet.ibm.com>
On 01/19/2011 10:39 AM, Srivatsa Vaddagiri wrote:
> I have tested quite extensively with booting a 16-vcpu guest (on a 16-pcpu host)
> and running kernel compine (with 32-threads). Without this patch, I had
> difficulty booting/shutting-down successfully (it would hang mid-way).
Sounds good. But I like to test with "make -j 100-200" to really give
things a workout.
J
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Jeremy Fitzhardinge @ 2011-01-19 18:53 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm,
Linux Kernel Mailing List, vatsa, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <1295457672.28776.144.camel@laptop>
On 01/19/2011 09:21 AM, Peter Zijlstra wrote:
> On Wed, 2011-01-19 at 22:42 +0530, Srivatsa Vaddagiri wrote:
>> Add two hypercalls to KVM hypervisor to support pv-ticketlocks.
>>
>> KVM_HC_WAIT_FOR_KICK blocks the calling vcpu until another vcpu kicks it or it
>> is woken up because of an event like interrupt.
>>
>> KVM_HC_KICK_CPU allows the calling vcpu to kick another vcpu.
>>
>> The presence of these hypercalls is indicated to guest via
>> KVM_FEATURE_WAIT_FOR_KICK/KVM_CAP_WAIT_FOR_KICK. Qemu needs a corresponding
>> patch to pass up the presence of this feature to guest via cpuid. Patch to qemu
>> will be sent separately.
>
> I didn't really read the patch, and I totally forgot everything from
> when I looked at the Xen series, but does the Xen/KVM hypercall
> interface for this include the vcpu to await the kick from?
>
> My guess is not, since the ticket locks used don't know who the owner
> is, which is of course, sad. There are FIFO spinlock implementations
> that can do this though.. although I think they all have a bigger memory
> footprint.
At least in the Xen code, a current owner isn't very useful, because we
need the current owner to kick the *next* owner to life at release time,
which we can't do without some structure recording which ticket belongs
to which cpu.
(A reminder: the big problem with ticket locks is not with the current
owner getting preempted, but making sure the next VCPU gets scheduled
quickly when the current one releases; without that all the waiting
VCPUs burn the timeslices until the VCPU scheduler gets around to
scheduling the actual next in line.)
At present, the code needs to scan an array of percpu "I am waiting on
lock X with ticket Y" structures to work out who's next. The search is
somewhat optimised by keeping a cpuset of which CPUs are actually
blocked on spinlocks, but its still going to scale badly with lots of CPUs.
I haven't thought of a good way to improve on this; an obvious approach
is to just add a pointer to the spinlock and hang an explicit linked
list off it, but that's incompatible with wanting to avoid expanding the
lock.
You could have a table of auxiliary per-lock data hashed on the lock
address, but its not clear to me that its an improvement on the array
approach, especially given the synchronization issues of keeping that
structure up to date (do we have a generic lockless hashtable
implementation?). But perhaps its one of those things that makes sense
at larger scales.
> The reason for wanting this should be clear I guess, it allows PI.
Well, if we can expand the spinlock to include an owner, then all this
becomes moot...
J
^ permalink raw reply
* Re: [PATCH 13/14] x86/ticketlock: add slowpath logic
From: Srivatsa Vaddagiri @ 2011-01-19 18:39 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <4D372DEA.1060004@goop.org>
On Wed, Jan 19, 2011 at 10:31:06AM -0800, Jeremy Fitzhardinge wrote:
> I think you're probably right; when I last tested this code, it was
> hanging in at about the rate this kind of race would cause. And in my
> previous analysis of similar schemes (the current pv spinlock code), it
> was always correct to do the "slowpath" test after doing do the unlock.
>
> *But* I haven't yet had the chance to specifically go through and
> analyse your patch in detail to make sure there's nothing else going on,
> so take this as provisional.
>
> How much testing have you given it?
I have tested quite extensively with booting a 16-vcpu guest (on a 16-pcpu host)
and running kernel compine (with 32-threads). Without this patch, I had
difficulty booting/shutting-down successfully (it would hang mid-way).
- vatsa
^ permalink raw reply
* Re: [PATCH 13/14] x86/ticketlock: add slowpath logic
From: Jeremy Fitzhardinge @ 2011-01-19 18:31 UTC (permalink / raw)
To: vatsa
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110117152222.GA19233@linux.vnet.ibm.com>
On 01/17/2011 07:22 AM, Srivatsa Vaddagiri wrote:
> On Tue, Nov 16, 2010 at 01:08:44PM -0800, Jeremy Fitzhardinge wrote:
>> From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
>>
>> Maintain a flag in both LSBs of the ticket lock which indicates whether
>> anyone is in the lock slowpath and may need kicking when the current
>> holder unlocks. The flags are set when the first locker enters
>> the slowpath, and cleared when unlocking to an empty queue.
>>
>> In the specific implementation of lock_spinning(), make sure to set
>> the slowpath flags on the lock just before blocking. We must do
>> this before the last-chance pickup test to prevent a deadlock
>> with the unlocker:
>>
>> Unlocker Locker
>> test for lock pickup
>> -> fail
>> test slowpath + unlock
>> -> false
>> set slowpath flags
>> block
>>
>> Whereas this works in any ordering:
>>
>> Unlocker Locker
>> set slowpath flags
>> test for lock pickup
>> -> fail
>> block
>> test slowpath + unlock
>> -> true, kick
> I think this is still racy ..
>
> Unlocker Locker
>
>
> test slowpath
> -> false
>
> set slowpath flag
> test for lock pickup
> -> fail
> block
>
>
> unlock
>
> unlock needs to happen first before testing slowpath? I have made that change
> for my KVM guest and it seems to be working well with that change .. Will
> cleanup and post my patches shortly
I think you're probably right; when I last tested this code, it was
hanging in at about the rate this kind of race would cause. And in my
previous analysis of similar schemes (the current pv spinlock code), it
was always correct to do the "slowpath" test after doing do the unlock.
*But* I haven't yet had the chance to specifically go through and
analyse your patch in detail to make sure there's nothing else going on,
so take this as provisional.
How much testing have you given it?
Thanks,
J
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Srivatsa Vaddagiri @ 2011-01-19 18:29 UTC (permalink / raw)
To: Peter Zijlstra
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <1295457672.28776.144.camel@laptop>
On Wed, Jan 19, 2011 at 06:21:12PM +0100, Peter Zijlstra wrote:
> I didn't really read the patch, and I totally forgot everything from
> when I looked at the Xen series, but does the Xen/KVM hypercall
> interface for this include the vcpu to await the kick from?
No not yet, for reasons you mention below.
> My guess is not, since the ticket locks used don't know who the owner
> is, which is of course, sad. There are FIFO spinlock implementations
> that can do this though.. although I think they all have a bigger memory
> footprint.
>
> The reason for wanting this should be clear I guess, it allows PI.
Yeah - that would be good to experiment with.
- vatsa
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Peter Zijlstra @ 2011-01-19 17:50 UTC (permalink / raw)
To: vatsa
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119172306.GA2863@linux.vnet.ibm.com>
On Wed, 2011-01-19 at 22:53 +0530, Srivatsa Vaddagiri wrote:
> On Wed, Jan 19, 2011 at 10:42:39PM +0530, Srivatsa Vaddagiri wrote:
> > Add two hypercalls to KVM hypervisor to support pv-ticketlocks.
> >
> > KVM_HC_WAIT_FOR_KICK blocks the calling vcpu until another vcpu kicks it or it
> > is woken up because of an event like interrupt.
>
> One possibility is to extend this hypercall to do a directed yield as well,
> which needs some more thought. Another issue that needs to be resolved with
> pv-ticketlocks is the impact on intra-VM fairness. A guest experiencing heavy
> contention can keep yielding cpu, allowing other VMs to get more time than they
> deserve.
No, yield sucks, if you know the owner you can do actual PI.
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Srivatsa Vaddagiri @ 2011-01-19 17:23 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119171239.GB726@linux.vnet.ibm.com>
On Wed, Jan 19, 2011 at 10:42:39PM +0530, Srivatsa Vaddagiri wrote:
> Add two hypercalls to KVM hypervisor to support pv-ticketlocks.
>
> KVM_HC_WAIT_FOR_KICK blocks the calling vcpu until another vcpu kicks it or it
> is woken up because of an event like interrupt.
One possibility is to extend this hypercall to do a directed yield as well,
which needs some more thought. Another issue that needs to be resolved with
pv-ticketlocks is the impact on intra-VM fairness. A guest experiencing heavy
contention can keep yielding cpu, allowing other VMs to get more time than they
deserve.
- vatsa
^ permalink raw reply
* Re: [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Peter Zijlstra @ 2011-01-19 17:21 UTC (permalink / raw)
To: vatsa
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119171239.GB726@linux.vnet.ibm.com>
On Wed, 2011-01-19 at 22:42 +0530, Srivatsa Vaddagiri wrote:
> Add two hypercalls to KVM hypervisor to support pv-ticketlocks.
>
> KVM_HC_WAIT_FOR_KICK blocks the calling vcpu until another vcpu kicks it or it
> is woken up because of an event like interrupt.
>
> KVM_HC_KICK_CPU allows the calling vcpu to kick another vcpu.
>
> The presence of these hypercalls is indicated to guest via
> KVM_FEATURE_WAIT_FOR_KICK/KVM_CAP_WAIT_FOR_KICK. Qemu needs a corresponding
> patch to pass up the presence of this feature to guest via cpuid. Patch to qemu
> will be sent separately.
I didn't really read the patch, and I totally forgot everything from
when I looked at the Xen series, but does the Xen/KVM hypercall
interface for this include the vcpu to await the kick from?
My guess is not, since the ticket locks used don't know who the owner
is, which is of course, sad. There are FIFO spinlock implementations
that can do this though.. although I think they all have a bigger memory
footprint.
The reason for wanting this should be clear I guess, it allows PI.
^ permalink raw reply
* [PATCH 3/3] kvm guest : Add support for pv-ticketlocks
From: Srivatsa Vaddagiri @ 2011-01-19 17:17 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119164432.GA30669@linux.vnet.ibm.com>
This patch extends Linux guests running on KVM hypervisor to support
pv-ticketlocks. Very early during bootup, paravirtualied KVM guest detects if
the hypervisor has required feature (KVM_FEATURE_WAIT_FOR_KICK) to support
pv-ticketlocks. If so, support for pv-ticketlocks is registered via pv_lock_ops.
Signed-off-by: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Signed-off-by: Suzuki Poulose <suzuki@in.ibm.com>
---
arch/x86/Kconfig | 9 +
arch/x86/include/asm/kvm_para.h | 8 +
arch/x86/kernel/head64.c | 3
arch/x86/kernel/kvm.c | 208 ++++++++++++++++++++++++++++++++++++++++
4 files changed, 228 insertions(+)
Index: linux-2.6.37/arch/x86/Kconfig
===================================================================
--- linux-2.6.37.orig/arch/x86/Kconfig
+++ linux-2.6.37/arch/x86/Kconfig
@@ -508,6 +508,15 @@ config KVM_GUEST
This option enables various optimizations for running under the KVM
hypervisor.
+config KVM_DEBUG_FS
+ bool "Enable debug information for KVM Guests in debugfs"
+ depends on KVM_GUEST
+ default n
+ ---help---
+ This option enables collection of various statistics for KVM guest.
+ Statistics are displayed in debugfs filesystem. Enabling this option
+ may incur significant overhead.
+
source "arch/x86/lguest/Kconfig"
config PARAVIRT
Index: linux-2.6.37/arch/x86/include/asm/kvm_para.h
===================================================================
--- linux-2.6.37.orig/arch/x86/include/asm/kvm_para.h
+++ linux-2.6.37/arch/x86/include/asm/kvm_para.h
@@ -162,8 +162,16 @@ static inline unsigned int kvm_arch_para
#ifdef CONFIG_KVM_GUEST
void __init kvm_guest_init(void);
+
+#ifdef CONFIG_PARAVIRT_SPINLOCKS
+void __init kvm_guest_early_init(void);
+#else
+#define kvm_guest_early_init() do { } while (0)
+#endif
+
#else
#define kvm_guest_init() do { } while (0)
+#define kvm_guest_early_init() do { } while (0)
#endif
#endif /* __KERNEL__ */
Index: linux-2.6.37/arch/x86/kernel/head64.c
===================================================================
--- linux-2.6.37.orig/arch/x86/kernel/head64.c
+++ linux-2.6.37/arch/x86/kernel/head64.c
@@ -13,6 +13,7 @@
#include <linux/start_kernel.h>
#include <linux/io.h>
#include <linux/memblock.h>
+#include <linux/kvm_para.h>
#include <asm/processor.h>
#include <asm/proto.h>
@@ -118,6 +119,8 @@ void __init x86_64_start_reservations(ch
reserve_ebda_region();
+ kvm_guest_early_init();
+
/*
* At this point everything still needed from the boot loader
* or BIOS or kernel text should be early reserved or marked not
Index: linux-2.6.37/arch/x86/kernel/kvm.c
===================================================================
--- linux-2.6.37.orig/arch/x86/kernel/kvm.c
+++ linux-2.6.37/arch/x86/kernel/kvm.c
@@ -238,3 +238,211 @@ void __init kvm_guest_init(void)
paravirt_ops_setup();
}
+
+#ifdef CONFIG_PARAVIRT_SPINLOCKS
+
+#ifdef CONFIG_KVM_DEBUG_FS
+
+#include <linux/debugfs.h>
+#include <linux/sched.h>
+
+static struct kvm_spinlock_stats
+{
+ u32 taken_slow;
+ u32 taken_slow_pickup;
+
+ u32 released_slow;
+ u32 released_slow_kicked;
+
+#define HISTO_BUCKETS 30
+ u32 histo_spin_blocked[HISTO_BUCKETS+1];
+
+ u64 time_blocked;
+} spinlock_stats;
+
+static u8 zero_stats;
+
+static inline void check_zero(void)
+{
+ if (unlikely(zero_stats)) {
+ memset(&spinlock_stats, 0, sizeof(spinlock_stats));
+ zero_stats = 0;
+ }
+}
+
+#define ADD_STATS(elem, val) \
+ do { check_zero(); spinlock_stats.elem += (val); } while (0)
+
+static inline u64 spin_time_start(void)
+{
+ return sched_clock();
+}
+
+static void __spin_time_accum(u64 delta, u32 *array)
+{
+ unsigned index = ilog2(delta);
+
+ check_zero();
+
+ if (index < HISTO_BUCKETS)
+ array[index]++;
+ else
+ array[HISTO_BUCKETS]++;
+}
+
+static inline void spin_time_accum_blocked(u64 start)
+{
+ u32 delta = sched_clock() - start;
+
+ __spin_time_accum(delta, spinlock_stats.histo_spin_blocked);
+ spinlock_stats.time_blocked += delta;
+}
+
+static struct dentry *d_spin_debug;
+static struct dentry *d_kvm_debug;
+
+struct dentry *kvm_init_debugfs(void)
+{
+ d_kvm_debug = debugfs_create_dir("kvm", NULL);
+ if (!d_kvm_debug)
+ printk(KERN_WARNING "Could not create 'kvm' debugfs directory\n");
+
+ return d_kvm_debug;
+}
+
+static int __init kvm_spinlock_debugfs(void)
+{
+ struct dentry *d_kvm = kvm_init_debugfs();
+
+ if (d_kvm == NULL)
+ return -ENOMEM;
+
+ d_spin_debug = debugfs_create_dir("spinlocks", d_kvm);
+
+ debugfs_create_u8("zero_stats", 0644, d_spin_debug, &zero_stats);
+
+ debugfs_create_u32("taken_slow", 0444, d_spin_debug,
+ &spinlock_stats.taken_slow);
+ debugfs_create_u32("taken_slow_pickup", 0444, d_spin_debug,
+ &spinlock_stats.taken_slow_pickup);
+
+ debugfs_create_u32("released_slow", 0444, d_spin_debug,
+ &spinlock_stats.released_slow);
+ debugfs_create_u32("released_slow_kicked", 0444, d_spin_debug,
+ &spinlock_stats.released_slow_kicked);
+
+ debugfs_create_u64("time_blocked", 0444, d_spin_debug,
+ &spinlock_stats.time_blocked);
+
+ debugfs_create_u32_array("histo_blocked", 0444, d_spin_debug,
+ spinlock_stats.histo_spin_blocked, HISTO_BUCKETS + 1);
+
+ return 0;
+}
+fs_initcall(kvm_spinlock_debugfs);
+#else /* !CONFIG_KVM_DEBUG_FS */
+#define TIMEOUT (1 << 10)
+#define ADD_STATS(elem, val) do { (void)(val); } while (0)
+
+static inline u64 spin_time_start(void)
+{
+ return 0;
+}
+
+static inline void spin_time_accum_blocked(u64 start)
+{
+}
+#endif /* CONFIG_KVM_DEBUG_FS */
+
+struct kvm_lock_waiting {
+ struct arch_spinlock *lock;
+ __ticket_t want;
+};
+
+/* cpus 'waiting' on a spinlock to become available */
+static cpumask_t waiting_cpus;
+
+/* Track spinlock on which a cpu is waiting */
+static DEFINE_PER_CPU(struct kvm_lock_waiting, lock_waiting);
+
+static inline void kvm_wait_for_kick(void)
+{
+ kvm_hypercall0(KVM_HC_WAIT_FOR_KICK);
+}
+
+static void kvm_lock_spinning(struct arch_spinlock *lock, unsigned want)
+{
+ struct kvm_lock_waiting *w = &__get_cpu_var(lock_waiting);
+ int cpu = smp_processor_id();
+ u64 start;
+
+ start = spin_time_start();
+
+ w->want = want;
+ w->lock = lock;
+
+ ADD_STATS(taken_slow, 1);
+
+ cpumask_set_cpu(cpu, &waiting_cpus);
+
+ /* Mark entry to slowpath before doing the pickup test to make
+ sure we don't deadlock with an unlocker. */
+ __ticket_enter_slowpath(lock);
+
+ /* check again make sure it didn't become free while
+ we weren't looking */
+ if (ACCESS_ONCE(lock->tickets.head) == want) {
+ ADD_STATS(taken_slow_pickup, 1);
+ goto out;
+ }
+
+ kvm_wait_for_kick();
+
+out:
+ cpumask_clear_cpu(cpu, &waiting_cpus);
+ w->lock = NULL;
+ spin_time_accum_blocked(start);
+}
+PV_CALLEE_SAVE_REGS_THUNK(kvm_lock_spinning);
+
+/* Kick a cpu */
+static inline void kvm_kick_cpu(int cpu)
+{
+ kvm_hypercall1(KVM_HC_KICK_CPU, cpu);
+}
+
+/* Kick vcpu waiting on @lock->head to reach value @ticket */
+static void kvm_unlock_kick(struct arch_spinlock *lock, unsigned ticket)
+{
+ int cpu;
+
+ ADD_STATS(released_slow, 1);
+
+ for_each_cpu(cpu, &waiting_cpus) {
+ const struct kvm_lock_waiting *w = &per_cpu(lock_waiting, cpu);
+ if (w->lock == lock && w->want == ticket) {
+ ADD_STATS(released_slow_kicked, 1);
+ kvm_kick_cpu(cpu);
+ break;
+ }
+ }
+}
+
+/*
+ * Setup pv_lock_ops to exploit KVM_FEATURE_WAIT_FOR_KICK if present.
+ * This needs to be setup really early in boot, before the first call to
+ * spinlock is issued!
+ */
+void __init kvm_guest_early_init(void)
+{
+ if (!kvm_para_available())
+ return;
+
+ /* Does host kernel support KVM_FEATURE_WAIT_FOR_KICK? */
+ if (!kvm_para_has_feature(KVM_FEATURE_WAIT_FOR_KICK))
+ return;
+
+ pv_lock_ops.lock_spinning = PV_CALLEE_SAVE(kvm_lock_spinning);
+ pv_lock_ops.unlock_kick = kvm_unlock_kick;
+}
+#endif /* CONFIG_PARAVIRT_SPINLOCKS */
^ permalink raw reply
* [PATCH 2/3] kvm hypervisor : Add hypercalls to support pv-ticketlock
From: Srivatsa Vaddagiri @ 2011-01-19 17:12 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119164432.GA30669@linux.vnet.ibm.com>
Add two hypercalls to KVM hypervisor to support pv-ticketlocks.
KVM_HC_WAIT_FOR_KICK blocks the calling vcpu until another vcpu kicks it or it
is woken up because of an event like interrupt.
KVM_HC_KICK_CPU allows the calling vcpu to kick another vcpu.
The presence of these hypercalls is indicated to guest via
KVM_FEATURE_WAIT_FOR_KICK/KVM_CAP_WAIT_FOR_KICK. Qemu needs a corresponding
patch to pass up the presence of this feature to guest via cpuid. Patch to qemu
will be sent separately.
Signed-off-by: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Signed-off-by: Suzuki Poulose <suzuki@in.ibm.com>
---
arch/x86/include/asm/kvm_para.h | 2 +
arch/x86/kvm/x86.c | 69 +++++++++++++++++++++++++++++++++++++++-
include/linux/kvm.h | 1
include/linux/kvm_host.h | 1
include/linux/kvm_para.h | 2 +
virt/kvm/kvm_main.c | 1
6 files changed, 75 insertions(+), 1 deletion(-)
Index: linux-2.6.37/arch/x86/include/asm/kvm_para.h
===================================================================
--- linux-2.6.37.orig/arch/x86/include/asm/kvm_para.h
+++ linux-2.6.37/arch/x86/include/asm/kvm_para.h
@@ -16,6 +16,8 @@
#define KVM_FEATURE_CLOCKSOURCE 0
#define KVM_FEATURE_NOP_IO_DELAY 1
#define KVM_FEATURE_MMU_OP 2
+#define KVM_FEATURE_WAIT_FOR_KICK 4
+
/* This indicates that the new set of kvmclock msrs
* are available. The use of 0x11 and 0x12 is deprecated
*/
Index: linux-2.6.37/arch/x86/kvm/x86.c
===================================================================
--- linux-2.6.37.orig/arch/x86/kvm/x86.c
+++ linux-2.6.37/arch/x86/kvm/x86.c
@@ -1922,6 +1922,7 @@ int kvm_dev_ioctl_check_extension(long e
case KVM_CAP_DEBUGREGS:
case KVM_CAP_X86_ROBUST_SINGLESTEP:
case KVM_CAP_XSAVE:
+ case KVM_CAP_WAIT_FOR_KICK:
r = 1;
break;
case KVM_CAP_COALESCED_MMIO:
@@ -2340,7 +2341,8 @@ static void do_cpuid_ent(struct kvm_cpui
entry->eax = (1 << KVM_FEATURE_CLOCKSOURCE) |
(1 << KVM_FEATURE_NOP_IO_DELAY) |
(1 << KVM_FEATURE_CLOCKSOURCE2) |
- (1 << KVM_FEATURE_CLOCKSOURCE_STABLE_BIT);
+ (1 << KVM_FEATURE_CLOCKSOURCE_STABLE_BIT) |
+ (1 << KVM_FEATURE_WAIT_FOR_KICK);
entry->ebx = 0;
entry->ecx = 0;
entry->edx = 0;
@@ -4766,6 +4768,63 @@ int kvm_hv_hypercall(struct kvm_vcpu *vc
return 1;
}
+/*
+ * kvm_pv_wait_for_kick_op : Block until kicked by either a KVM_HC_KICK_CPU
+ * hypercall or a event like interrupt.
+ *
+ * @vcpu : vcpu which is blocking.
+ */
+static void kvm_pv_wait_for_kick_op(struct kvm_vcpu *vcpu)
+{
+ DEFINE_WAIT(wait);
+
+ /*
+ * Blocking on vcpu->wq allows us to wake up sooner if required to
+ * service pending events (like interrupts).
+ *
+ * Also set state to TASK_INTERRUPTIBLE before checking vcpu->kicked to
+ * avoid racing with kvm_pv_kick_cpu_op().
+ */
+ prepare_to_wait(&vcpu->wq, &wait, TASK_INTERRUPTIBLE);
+
+ do {
+ /*
+ * Somebody has already tried kicking us. Acknowledge that
+ * and terminate the wait.
+ */
+ if (vcpu->kicked) {
+ vcpu->kicked = 0;
+ break;
+ }
+
+ /* Let's wait for either KVM_HC_KICK_CPU or someother event
+ * to wake us up.
+ */
+
+ srcu_read_unlock(&vcpu->kvm->srcu, vcpu->srcu_idx);
+ schedule();
+ vcpu->srcu_idx = srcu_read_lock(&vcpu->kvm->srcu);
+
+ } while (0);
+
+ finish_wait(&vcpu->wq, &wait);
+}
+
+/*
+ * kvm_pv_kick_cpu_op: Kick a vcpu.
+ *
+ * @cpu - vcpu to be kicked.
+ */
+static void kvm_pv_kick_cpu_op(struct kvm *kvm, int cpu)
+{
+ struct kvm_vcpu *vcpu = kvm_get_vcpu(kvm, cpu);
+
+ if (vcpu) {
+ vcpu->kicked = 1;
+ wake_up_interruptible(&vcpu->wq);
+ }
+}
+
int kvm_emulate_hypercall(struct kvm_vcpu *vcpu)
{
unsigned long nr, a0, a1, a2, a3, ret;
@@ -4802,6 +4861,14 @@ int kvm_emulate_hypercall(struct kvm_vcp
case KVM_HC_MMU_OP:
r = kvm_pv_mmu_op(vcpu, a0, hc_gpa(vcpu, a1, a2), &ret);
break;
+ case KVM_HC_WAIT_FOR_KICK:
+ kvm_pv_wait_for_kick_op(vcpu);
+ ret = 0;
+ break;
+ case KVM_HC_KICK_CPU:
+ kvm_pv_kick_cpu_op(vcpu->kvm, a0);
+ ret = 0;
+ break;
default:
ret = -KVM_ENOSYS;
break;
Index: linux-2.6.37/include/linux/kvm.h
===================================================================
--- linux-2.6.37.orig/include/linux/kvm.h
+++ linux-2.6.37/include/linux/kvm.h
@@ -540,6 +540,7 @@ struct kvm_ppc_pvinfo {
#endif
#define KVM_CAP_PPC_GET_PVINFO 57
#define KVM_CAP_PPC_IRQ_LEVEL 58
+#define KVM_CAP_WAIT_FOR_KICK 59
#ifdef KVM_CAP_IRQ_ROUTING
Index: linux-2.6.37/include/linux/kvm_host.h
===================================================================
--- linux-2.6.37.orig/include/linux/kvm_host.h
+++ linux-2.6.37/include/linux/kvm_host.h
@@ -105,6 +105,7 @@ struct kvm_vcpu {
#endif
struct kvm_vcpu_arch arch;
+ int kicked;
};
/*
Index: linux-2.6.37/include/linux/kvm_para.h
===================================================================
--- linux-2.6.37.orig/include/linux/kvm_para.h
+++ linux-2.6.37/include/linux/kvm_para.h
@@ -19,6 +19,8 @@
#define KVM_HC_MMU_OP 2
#define KVM_HC_FEATURES 3
#define KVM_HC_PPC_MAP_MAGIC_PAGE 4
+#define KVM_HC_WAIT_FOR_KICK 5
+#define KVM_HC_KICK_CPU 6
/*
* hypercalls use architecture specific
Index: linux-2.6.37/virt/kvm/kvm_main.c
===================================================================
--- linux-2.6.37.orig/virt/kvm/kvm_main.c
+++ linux-2.6.37/virt/kvm/kvm_main.c
@@ -185,6 +185,7 @@ int kvm_vcpu_init(struct kvm_vcpu *vcpu,
vcpu->cpu = -1;
vcpu->kvm = kvm;
vcpu->vcpu_id = id;
+ vcpu->kicked = 0;
init_waitqueue_head(&vcpu->wq);
page = alloc_page(GFP_KERNEL | __GFP_ZERO);
^ permalink raw reply
* [PATCH 1/3] debugfs: Add support to print u32 array
From: Srivatsa Vaddagiri @ 2011-01-19 17:07 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110119164432.GA30669@linux.vnet.ibm.com>
Add debugfs support to print u32-arrays.
Most of this comes from Xen-hypervisor sources, which has been refactored to
make the code common for other users as well.
Signed-off-by: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
Signed-off-by: Suzuki Poulose <suzuki@in.ibm.com>
---
arch/x86/xen/debugfs.c | 104 --------------------------------------------
arch/x86/xen/debugfs.h | 4 -
arch/x86/xen/mmu.c | 2
arch/x86/xen/multicalls.c | 6 +-
arch/x86/xen/spinlock.c | 2
fs/debugfs/file.c | 108 ++++++++++++++++++++++++++++++++++++++++++++++
include/linux/debugfs.h | 11 ++++
7 files changed, 124 insertions(+), 113 deletions(-)
Index: linux-2.6.37/arch/x86/xen/debugfs.c
===================================================================
--- linux-2.6.37.orig/arch/x86/xen/debugfs.c
+++ linux-2.6.37/arch/x86/xen/debugfs.c
@@ -19,107 +19,3 @@ struct dentry * __init xen_init_debugfs(
return d_xen_debug;
}
-struct array_data
-{
- void *array;
- unsigned elements;
-};
-
-static int u32_array_open(struct inode *inode, struct file *file)
-{
- file->private_data = NULL;
- return nonseekable_open(inode, file);
-}
-
-static size_t format_array(char *buf, size_t bufsize, const char *fmt,
- u32 *array, unsigned array_size)
-{
- size_t ret = 0;
- unsigned i;
-
- for(i = 0; i < array_size; i++) {
- size_t len;
-
- len = snprintf(buf, bufsize, fmt, array[i]);
- len++; /* ' ' or '\n' */
- ret += len;
-
- if (buf) {
- buf += len;
- bufsize -= len;
- buf[-1] = (i == array_size-1) ? '\n' : ' ';
- }
- }
-
- ret++; /* \0 */
- if (buf)
- *buf = '\0';
-
- return ret;
-}
-
-static char *format_array_alloc(const char *fmt, u32 *array, unsigned array_size)
-{
- size_t len = format_array(NULL, 0, fmt, array, array_size);
- char *ret;
-
- ret = kmalloc(len, GFP_KERNEL);
- if (ret == NULL)
- return NULL;
-
- format_array(ret, len, fmt, array, array_size);
- return ret;
-}
-
-static ssize_t u32_array_read(struct file *file, char __user *buf, size_t len,
- loff_t *ppos)
-{
- struct inode *inode = file->f_path.dentry->d_inode;
- struct array_data *data = inode->i_private;
- size_t size;
-
- if (*ppos == 0) {
- if (file->private_data) {
- kfree(file->private_data);
- file->private_data = NULL;
- }
-
- file->private_data = format_array_alloc("%u", data->array, data->elements);
- }
-
- size = 0;
- if (file->private_data)
- size = strlen(file->private_data);
-
- return simple_read_from_buffer(buf, len, ppos, file->private_data, size);
-}
-
-static int xen_array_release(struct inode *inode, struct file *file)
-{
- kfree(file->private_data);
-
- return 0;
-}
-
-static const struct file_operations u32_array_fops = {
- .owner = THIS_MODULE,
- .open = u32_array_open,
- .release= xen_array_release,
- .read = u32_array_read,
- .llseek = no_llseek,
-};
-
-struct dentry *xen_debugfs_create_u32_array(const char *name, mode_t mode,
- struct dentry *parent,
- u32 *array, unsigned elements)
-{
- struct array_data *data = kmalloc(sizeof(*data), GFP_KERNEL);
-
- if (data == NULL)
- return NULL;
-
- data->array = array;
- data->elements = elements;
-
- return debugfs_create_file(name, mode, parent, data, &u32_array_fops);
-}
Index: linux-2.6.37/arch/x86/xen/debugfs.h
===================================================================
--- linux-2.6.37.orig/arch/x86/xen/debugfs.h
+++ linux-2.6.37/arch/x86/xen/debugfs.h
@@ -3,8 +3,4 @@
struct dentry * __init xen_init_debugfs(void);
-struct dentry *xen_debugfs_create_u32_array(const char *name, mode_t mode,
- struct dentry *parent,
- u32 *array, unsigned elements);
-
#endif /* _XEN_DEBUGFS_H */
Index: linux-2.6.37/arch/x86/xen/mmu.c
===================================================================
--- linux-2.6.37.orig/arch/x86/xen/mmu.c
+++ linux-2.6.37/arch/x86/xen/mmu.c
@@ -2757,7 +2757,7 @@ static int __init xen_mmu_debugfs(void)
debugfs_create_u32("mmu_update", 0444, d_mmu_debug, &mmu_stats.mmu_update);
debugfs_create_u32("mmu_update_extended", 0444, d_mmu_debug,
&mmu_stats.mmu_update_extended);
- xen_debugfs_create_u32_array("mmu_update_histo", 0444, d_mmu_debug,
+ debugfs_create_u32_array("mmu_update_histo", 0444, d_mmu_debug,
mmu_stats.mmu_update_histo, 20);
debugfs_create_u32("set_pte_at", 0444, d_mmu_debug, &mmu_stats.set_pte_at);
Index: linux-2.6.37/arch/x86/xen/multicalls.c
===================================================================
--- linux-2.6.37.orig/arch/x86/xen/multicalls.c
+++ linux-2.6.37/arch/x86/xen/multicalls.c
@@ -269,11 +269,11 @@ static int __init xen_mc_debugfs(void)
debugfs_create_u32("hypercalls", 0444, d_mc_debug, &mc_stats.hypercalls);
debugfs_create_u32("arg_total", 0444, d_mc_debug, &mc_stats.arg_total);
- xen_debugfs_create_u32_array("batch_histo", 0444, d_mc_debug,
+ debugfs_create_u32_array("batch_histo", 0444, d_mc_debug,
mc_stats.histo, MC_BATCH);
- xen_debugfs_create_u32_array("hypercall_histo", 0444, d_mc_debug,
+ debugfs_create_u32_array("hypercall_histo", 0444, d_mc_debug,
mc_stats.histo_hypercalls, NHYPERCALLS);
- xen_debugfs_create_u32_array("flush_reasons", 0444, d_mc_debug,
+ debugfs_create_u32_array("flush_reasons", 0444, d_mc_debug,
mc_stats.flush, FL_N_REASONS);
return 0;
Index: linux-2.6.37/arch/x86/xen/spinlock.c
===================================================================
--- linux-2.6.37.orig/arch/x86/xen/spinlock.c
+++ linux-2.6.37/arch/x86/xen/spinlock.c
@@ -228,7 +228,7 @@ static int __init xen_spinlock_debugfs(v
debugfs_create_u64("time_blocked", 0444, d_spin_debug,
&spinlock_stats.time_blocked);
- xen_debugfs_create_u32_array("histo_blocked", 0444, d_spin_debug,
+ debugfs_create_u32_array("histo_blocked", 0444, d_spin_debug,
spinlock_stats.histo_spin_blocked, HISTO_BUCKETS + 1);
return 0;
Index: linux-2.6.37/fs/debugfs/file.c
===================================================================
--- linux-2.6.37.orig/fs/debugfs/file.c
+++ linux-2.6.37/fs/debugfs/file.c
@@ -18,6 +18,7 @@
#include <linux/pagemap.h>
#include <linux/namei.h>
#include <linux/debugfs.h>
+#include <linux/slab.h>
static ssize_t default_read_file(struct file *file, char __user *buf,
size_t count, loff_t *ppos)
@@ -534,3 +535,110 @@ struct dentry *debugfs_create_blob(const
return debugfs_create_file(name, mode, parent, blob, &fops_blob);
}
EXPORT_SYMBOL_GPL(debugfs_create_blob);
+
+struct array_data {
+ void *array;
+ unsigned elements;
+};
+
+static int u32_array_open(struct inode *inode, struct file *file)
+{
+ file->private_data = NULL;
+ return nonseekable_open(inode, file);
+}
+
+static size_t format_array(char *buf, size_t bufsize, const char *fmt,
+ u32 *array, unsigned array_size)
+{
+ size_t ret = 0;
+ unsigned i;
+
+ for (i = 0; i < array_size; i++) {
+ size_t len;
+
+ len = snprintf(buf, bufsize, fmt, array[i]);
+ len++; /* ' ' or '\n' */
+ ret += len;
+
+ if (buf) {
+ buf += len;
+ bufsize -= len;
+ buf[-1] = (i == array_size-1) ? '\n' : ' ';
+ }
+ }
+
+ ret++; /* \0 */
+ if (buf)
+ *buf = '\0';
+
+ return ret;
+}
+
+static char *
+format_array_alloc(const char *fmt, u32 *array, unsigned array_size)
+{
+ size_t len = format_array(NULL, 0, fmt, array, array_size);
+ char *ret;
+
+ ret = kmalloc(len, GFP_KERNEL);
+ if (ret == NULL)
+ return NULL;
+
+ format_array(ret, len, fmt, array, array_size);
+ return ret;
+}
+
+static ssize_t u32_array_read(struct file *file, char __user *buf, size_t len,
+ loff_t *ppos)
+{
+ struct inode *inode = file->f_path.dentry->d_inode;
+ struct array_data *data = inode->i_private;
+ size_t size;
+
+ if (*ppos == 0) {
+ if (file->private_data) {
+ kfree(file->private_data);
+ file->private_data = NULL;
+ }
+
+ file->private_data =
+ format_array_alloc("%u", data->array, data->elements);
+ }
+
+ size = 0;
+ if (file->private_data)
+ size = strlen(file->private_data);
+
+ return simple_read_from_buffer(buf, len, ppos,
+ file->private_data, size);
+}
+
+static int u32_array_release(struct inode *inode, struct file *file)
+{
+ kfree(file->private_data);
+
+ return 0;
+}
+
+static const struct file_operations u32_array_fops = {
+ .owner = THIS_MODULE,
+ .open = u32_array_open,
+ .release = u32_array_release,
+ .read = u32_array_read,
+ .llseek = no_llseek,
+};
+
+struct dentry *debugfs_create_u32_array(const char *name, mode_t mode,
+ struct dentry *parent,
+ u32 *array, unsigned elements)
+{
+ struct array_data *data = kmalloc(sizeof(*data), GFP_KERNEL);
+
+ if (data == NULL)
+ return NULL;
+
+ data->array = array;
+ data->elements = elements;
+
+ return debugfs_create_file(name, mode, parent, data, &u32_array_fops);
+}
Index: linux-2.6.37/include/linux/debugfs.h
===================================================================
--- linux-2.6.37.orig/include/linux/debugfs.h
+++ linux-2.6.37/include/linux/debugfs.h
@@ -74,6 +74,10 @@ struct dentry *debugfs_create_blob(const
struct dentry *parent,
struct debugfs_blob_wrapper *blob);
+struct dentry *debugfs_create_u32_array(const char *name, mode_t mode,
+ struct dentry *parent,
+ u32 *array, unsigned elements);
+
bool debugfs_initialized(void);
#else
@@ -187,6 +191,13 @@ static inline struct dentry *debugfs_cre
{
return ERR_PTR(-ENODEV);
}
+
+struct dentry *debugfs_create_u32_array(const char *name, mode_t mode,
+ struct dentry *parent,
+ u32 *array, unsigned elements)
+{
+ return ERR_PTR(-ENODEV);
+}
static inline bool debugfs_initialized(void)
{
^ permalink raw reply
* Re: [PATCH 00/14] PV ticket locks without expanding spinlock
From: Srivatsa Vaddagiri @ 2011-01-19 16:44 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, kvm, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <cover.1289940821.git.jeremy.fitzhardinge@citrix.com>
On Tue, Nov 16, 2010 at 01:08:31PM -0800, Jeremy Fitzhardinge wrote:
> From: Jeremy Fitzhardinge <jeremy.fitzhardinge@citrix.com>
>
> Hi all,
>
> This is a revised version of the pvticket lock series.
The 3-patch series to follow this email extends KVM-hypervisor and Linux guest
running on KVM-hypervisor to support pv-ticket spinlocks.
Two hypercalls are being introduced in KVM hypervisor, one that allows a
vcpu (spinning on a lock) to block and another that allows a vcpu to kick
another out of blocking state.
Patches are against 2.6.37 mainline kernel. I also don't yet have numbers
at this time to show benefit of pv-ticketlocks - I would think the benefit
should be similar across hypervisors (Xen and KVM in this case).
- vatsa
^ permalink raw reply
* Re: [PATCH 13/14] x86/ticketlock: add slowpath logic
From: Srivatsa Vaddagiri @ 2011-01-19 16:23 UTC (permalink / raw)
To: Jeremy Fitzhardinge
Cc: Xen-devel, Mathieu Desnoyers, Nick Piggin, Peter Zijlstra,
Linux Kernel Mailing List, Jan Beulich, Eric Dumazet,
Jeremy Fitzhardinge, suzuki, Avi Kivity, H. Peter Anvin,
Américo Wang, Linux Virtualization
In-Reply-To: <20110117152222.GA19233@linux.vnet.ibm.com>
On Mon, Jan 17, 2011 at 08:52:22PM +0530, Srivatsa Vaddagiri wrote:
> I think this is still racy ..
>
> Unlocker Locker
>
>
> test slowpath
> -> false
>
> set slowpath flag
> test for lock pickup
> -> fail
> block
>
>
> unlock
>
> unlock needs to happen first before testing slowpath? I have made that change
> for my KVM guest and it seems to be working well with that change .. Will
> cleanup and post my patches shortly
Patch below fixes the race described above. You can fold this to your patch
13/14 if you agree this is in right direction.
Signed-off-by: Srivatsa Vaddagiri <vatsa@linux.vnet.ibm.com>
---
arch/x86/include/asm/spinlock.h | 7 +++----
arch/x86/kernel/paravirt-spinlocks.c | 22 +++++-----------------
2 files changed, 8 insertions(+), 21 deletions(-)
Index: linux-2.6.37/arch/x86/include/asm/spinlock.h
===================================================================
--- linux-2.6.37.orig/arch/x86/include/asm/spinlock.h
+++ linux-2.6.37/arch/x86/include/asm/spinlock.h
@@ -55,7 +55,7 @@ static __always_inline void __ticket_unl
/* Only defined when CONFIG_PARAVIRT_SPINLOCKS defined, but may as
* well leave the prototype always visible. */
-extern void __ticket_unlock_release_slowpath(struct arch_spinlock *lock);
+extern void __ticket_unlock_slowpath(struct arch_spinlock *lock);
#ifdef CONFIG_PARAVIRT_SPINLOCKS
@@ -166,10 +166,9 @@ static __always_inline int arch_spin_try
static __always_inline void arch_spin_unlock(arch_spinlock_t *lock)
{
barrier(); /* prevent reordering out of locked region */
+ __ticket_unlock_release(lock);
if (unlikely(__ticket_in_slowpath(lock)))
- __ticket_unlock_release_slowpath(lock);
- else
- __ticket_unlock_release(lock);
+ __ticket_unlock_slowpath(lock);
barrier(); /* prevent reordering into locked region */
}
Index: linux-2.6.37/arch/x86/kernel/paravirt-spinlocks.c
===================================================================
--- linux-2.6.37.orig/arch/x86/kernel/paravirt-spinlocks.c
+++ linux-2.6.37/arch/x86/kernel/paravirt-spinlocks.c
@@ -22,33 +22,21 @@ EXPORT_SYMBOL(pv_lock_ops);
* bits. However, we need to be careful about this because someone
* may just be entering as we leave, and enter the slowpath.
*/
-void __ticket_unlock_release_slowpath(struct arch_spinlock *lock)
+void __ticket_unlock_slowpath(struct arch_spinlock *lock)
{
struct arch_spinlock old, new;
BUILD_BUG_ON(((__ticket_t)NR_CPUS) != NR_CPUS);
old = ACCESS_ONCE(*lock);
-
new = old;
- new.tickets.head += TICKET_LOCK_INC;
/* Clear the slowpath flag */
new.tickets.tail &= ~TICKET_SLOWPATH_FLAG;
+ if (new.tickets.head == new.tickets.tail)
+ cmpxchg(&lock->head_tail, old.head_tail, new.head_tail);
- /*
- * If there's currently people waiting or someone snuck in
- * since we read the lock above, then do a normal unlock and
- * kick. If we managed to unlock with no queued waiters, then
- * we can clear the slowpath flag.
- */
- if (new.tickets.head != new.tickets.tail ||
- cmpxchg(&lock->head_tail,
- old.head_tail, new.head_tail) != old.head_tail) {
- /* still people waiting */
- __ticket_unlock_release(lock);
- }
-
+ /* Wake up an appropriate waiter */
__ticket_unlock_kick(lock, new.tickets.head);
}
-EXPORT_SYMBOL(__ticket_unlock_release_slowpath);
+EXPORT_SYMBOL(__ticket_unlock_slowpath);
^ permalink raw reply
* Re: Flow Control and Port Mirroring Revisited
From: Simon Horman @ 2011-01-19 9:11 UTC (permalink / raw)
To: Michael S. Tsirkin
Cc: Rick Jones, Jesse Gross, Eric Dumazet, Rusty Russell,
virtualization, dev, virtualization, netdev, kvm
In-Reply-To: <20110118201333.GD18760@redhat.com>
On Tue, Jan 18, 2011 at 10:13:33PM +0200, Michael S. Tsirkin wrote:
> On Tue, Jan 18, 2011 at 11:41:22AM -0800, Rick Jones wrote:
> > >So it won't be all that simple to implement well, and before we try,
> > >I'd like to know whether there are applications that are helped
> > >by it. For example, we could try to measure latency at various
> > >pps and see whether the backpressure helps. netperf has -b, -w
> > >flags which might help these measurements.
> >
> > Those options are enabled when one adds --enable-burst to the
> > pre-compilation ./configure of netperf (one doesn't have to
> > recompile netserver). However, if one is also looking at latency
> > statistics via the -j option in the top-of-trunk, or simply at the
> > histogram with --enable-histogram on the ./configure and a verbosity
> > level of 2 (global -v 2) then one wants the very top of trunk
> > netperf from:
> >
> > http://www.netperf.org/svn/netperf2/trunk
> >
> > to get the recently added support for accurate (netperf level) RTT
> > measuremnts on burst-mode request/response tests.
> >
> > happy benchmarking,
> >
> > rick jones
Thanks Rick, that is really helpful.
> > PS - the enhanced latency statistics from -j are only available in
> > the "omni" version of the TCP_RR test. To get that add a
> > --enable-omni to the ./configure - and in this case both netperf and
> > netserver have to be recompiled.
>
>
> Is this TCP only? I would love to get latency data from UDP as well.
At a glance, -- -T UDP is what you are after.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox