Patchwork pci, e1000e: Add and use __pci_disable_link_state

login
register
mail settings
Submitter Yinghai Lu
Date May 8, 2011, 6:54 p.m.
Message ID <4DC6E6E8.9040306@kernel.org>
Download mbox | patch
Permalink /patch/94600/
State Awaiting Upstream
Delegated to: David Miller
Headers show

Comments

Yinghai Lu - May 8, 2011, 6:54 p.m.
Need to use it in _e1000e_disable_aspm.

Found lock up:

[ 2374.654557] kworker/32:1    D ffff881027f6b0f0     0  6075      2 0x00000000
[ 2374.654816]  ffff88503f099a68 0000000000000046 ffff88503f098000 0000000000004000
[ 2374.654837]  00000000001d1ec0 ffff88503f099fd8 00000000001d1ec0 ffff88503f099fd8
[ 2374.654860]  0000000000004000 00000000001d1ec0 ffff88503dcc8000 ffff88503f090000
[ 2374.654880] Call Trace:
[ 2374.654898]  [<ffffffff810b1302>] ? __lock_acquired+0x3a/0x224
[ 2374.654914]  [<ffffffff81c2b59c>] ? _raw_spin_unlock_irq+0x30/0x36
[ 2374.654925]  [<ffffffff810b069d>] ? trace_hardirqs_on_caller+0x1f/0x178
[ 2374.654936]  [<ffffffff81c2ab24>] rwsem_down_failed_common+0xd3/0x103
[ 2374.654945]  [<ffffffff810b158f>] ? __lock_contended+0x3a/0x2a2
[ 2374.654955]  [<ffffffff81c2ab7b>] rwsem_down_read_failed+0x12/0x14
[ 2374.654967]  [<ffffffff813371e4>] call_rwsem_down_read_failed+0x14/0x30
[ 2374.654981]  [<ffffffff8135df20>] ? pci_disable_link_state+0x5f/0xf5
[ 2374.654990]  [<ffffffff81c2a0e6>] ? down_read+0x7e/0x91
[ 2374.654999]  [<ffffffff8135df20>] ? pci_disable_link_state+0x5f/0xf5
[ 2374.655008]  [<ffffffff8135df20>] pci_disable_link_state+0x5f/0xf5
[ 2374.655024]  [<ffffffff81661796>] e1000e_disable_aspm+0x55/0x5a
[ 2374.655037]  [<ffffffff816677eb>] e1000_io_slot_reset+0x59/0xea
[ 2374.655048]  [<ffffffff8135fe0d>] ? report_mmio_enabled+0x5d/0x5d
[ 2374.655057]  [<ffffffff8135fe3b>] report_slot_reset+0x2e/0x5d
[ 2374.655072]  [<ffffffff8135369e>] pci_walk_bus+0x8a/0xb7
[ 2374.655081]  [<ffffffff8135fe0d>] ? report_mmio_enabled+0x5d/0x5d
[ 2374.655091]  [<ffffffff813603be>] broadcast_error_message+0xa4/0xb2
[ 2374.655101]  [<ffffffff81352c71>] ? pci_bus_read_config_dword+0x72/0x80
[ 2374.655110]  [<ffffffff813606df>] do_recovery+0x9e/0xf9
[ 2374.655120]  [<ffffffff81360786>] handle_error_source+0x4c/0x51
[ 2374.655129]  [<ffffffff81360974>] aer_isr_one_error+0x1e9/0x21a
[ 2374.655138]  [<ffffffff81360a6c>] aer_isr+0xc7/0xcc
[ 2374.655147]  [<ffffffff813609a5>] ? aer_isr_one_error+0x21a/0x21a
[ 2374.655159]  [<ffffffff81096d9f>] process_one_work+0x237/0x3ec
[ 2374.655168]  [<ffffffff81096d10>] ? process_one_work+0x1a8/0x3ec
[ 2374.655178]  [<ffffffff8109728d>] worker_thread+0x17c/0x240
[ 2374.655186]  [<ffffffff810b0803>] ? trace_hardirqs_on+0xd/0xf
[ 2374.655196]  [<ffffffff81097111>] ? manage_workers+0xab/0xab
[ 2374.655209]  [<ffffffff8109c8ed>] kthread+0xa0/0xa8
[ 2374.655223]  [<ffffffff81c332d4>] kernel_thread_helper+0x4/0x10
[ 2374.655232]  [<ffffffff81c2b880>] ? retint_restore_args+0xe/0xe
[ 2374.655243]  [<ffffffff8109c84d>] ? __init_kthread_worker+0x5b/0x5b
[ 2374.655252]  [<ffffffff81c332d0>] ? gs_change+0xb/0xb

when aer happens,
pci_walk_bus already have down_read(&pci_bus_sem)...
then report_slot_reset
        ==> e1000_io_slot_reset
                ==> e1000e_disable_aspm
                        ==> pci_disable_link_state...

We can not use pci_disable_link_state, and it will try to hold pci_bus_sem again.

Try to have __pci_disable_link_state that will not need to hold pci_bus_sem.

Signed-off-by: Yinghai Lu <yinghai@kernel.org>

---
 drivers/net/e1000e/netdev.c |    2 +-
 drivers/pci/pcie/aspm.c     |   16 +++++++++++++---
 include/linux/pci-aspm.h    |    1 +
 3 files changed, 15 insertions(+), 4 deletions(-)

--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Jesse Barnes - May 9, 2011, 9:35 p.m.
On Sun, 08 May 2011 11:54:32 -0700
Yinghai Lu <yinghai@kernel.org> wrote:

> 
> Need to use it in _e1000e_disable_aspm.
> when aer happens,
> pci_walk_bus already have down_read(&pci_bus_sem)...
> then report_slot_reset
>         ==> e1000_io_slot_reset
>                 ==> e1000e_disable_aspm
>                         ==> pci_disable_link_state...
> 
> We can not use pci_disable_link_state, and it will try to hold pci_bus_sem again.
> 
> Try to have __pci_disable_link_state that will not need to hold pci_bus_sem.

What about the other callers of e1000e_disable_aspm?  Do they already
have the lock held or is it just reset that needs the already locked
version?

> +extern void __pci_disable_link_state(struct pci_dev *pdev, int state);
>  extern void pcie_clear_aspm(void);
>  extern void pcie_no_aspm(void);
>  #else

pci_disable_link_state_locked would be more descriptive and would match
other usage in the kernel.

Thanks,
Yinghai Lu - May 11, 2011, 8:23 p.m.
On 05/09/2011 02:35 PM, Jesse Barnes wrote:
> On Sun, 08 May 2011 11:54:32 -0700
> Yinghai Lu <yinghai@kernel.org> wrote:
> 
>>
>> Need to use it in _e1000e_disable_aspm.
>> when aer happens,
>> pci_walk_bus already have down_read(&pci_bus_sem)...
>> then report_slot_reset
>>         ==> e1000_io_slot_reset
>>                 ==> e1000e_disable_aspm
>>                         ==> pci_disable_link_state...
>>
>> We can not use pci_disable_link_state, and it will try to hold pci_bus_sem again.
>>
>> Try to have __pci_disable_link_state that will not need to hold pci_bus_sem.
> 
> What about the other callers of e1000e_disable_aspm?  Do they already
> have the lock held or is it just reset that needs the already locked
> version?

yes. 

there is another version when aspm is not defined. and it does not use any lock. 

#ifdef CONFIG_PCIEASPM
static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
{
        pci_disable_link_state(pdev, state);
}
#else
static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
{
        int pos;
        u16 reg16;

        /*
         * Both device and parent should have the same ASPM setting.
         * Disable ASPM in downstream component first and then upstream.
         */
        pos = pci_pcie_cap(pdev);
        pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
        reg16 &= ~state;
        pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);

        if (!pdev->bus->self)
                return;

        pos = pci_pcie_cap(pdev->bus->self);
        pci_read_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, &reg16);
        reg16 &= ~state;
        pci_write_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, reg16);
}
#endif



> 
>> +extern void __pci_disable_link_state(struct pci_dev *pdev, int state);
>>  extern void pcie_clear_aspm(void);
>>  extern void pcie_no_aspm(void);
>>  #else
> 
> pci_disable_link_state_locked would be more descriptive and would match
> other usage in the kernel.

ok.

Thanks

Yinghai
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Jesse Barnes - May 12, 2011, 11:32 p.m.
On Wed, 11 May 2011 13:23:21 -0700
Yinghai Lu <yinghai@kernel.org> wrote:

> On 05/09/2011 02:35 PM, Jesse Barnes wrote:
> > On Sun, 08 May 2011 11:54:32 -0700
> > Yinghai Lu <yinghai@kernel.org> wrote:
> > 
> >>
> >> Need to use it in _e1000e_disable_aspm.
> >> when aer happens,
> >> pci_walk_bus already have down_read(&pci_bus_sem)...
> >> then report_slot_reset
> >>         ==> e1000_io_slot_reset
> >>                 ==> e1000e_disable_aspm
> >>                         ==> pci_disable_link_state...
> >>
> >> We can not use pci_disable_link_state, and it will try to hold pci_bus_sem again.
> >>
> >> Try to have __pci_disable_link_state that will not need to hold pci_bus_sem.
> > 
> > What about the other callers of e1000e_disable_aspm?  Do they already
> > have the lock held or is it just reset that needs the already locked
> > version?
> 
> yes. 
> 
> there is another version when aspm is not defined. and it does not use any lock. 
> 
> #ifdef CONFIG_PCIEASPM
> static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
> {
>         pci_disable_link_state(pdev, state);
> }
> #else
> static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
> {
>         int pos;
>         u16 reg16;
> 
>         /*
>          * Both device and parent should have the same ASPM setting.
>          * Disable ASPM in downstream component first and then upstream.
>          */
>         pos = pci_pcie_cap(pdev);
>         pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
>         reg16 &= ~state;
>         pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);
> 
>         if (!pdev->bus->self)
>                 return;
> 
>         pos = pci_pcie_cap(pdev->bus->self);
>         pci_read_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, &reg16);
>         reg16 &= ~state;
>         pci_write_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, reg16);
> }
> #endif

No, I mean __e1000e_disable_aspm is called from several spots:

*** drivers/net/e1000e/82571.c:
e1000_get_variants_82571[435]  e1000e_disable_aspm(adapter->pdev, PCIE_LINK_STATE_L0S);

*** drivers/net/e1000e/netdev.c:
e1000_change_mtu[5027]         e1000e_disable_aspm(adapter->pdev, PCIE_LINK_STATE_L1);
__e1000_resume[5402]           e1000e_disable_aspm(pdev, PCIE_LINK_STATE_L1);
e1000_io_slot_reset[5650]      e1000e_disable_aspm(pdev, PCIE_LINK_STATE_L1);
e1000_probe[5797]              e1000e_disable_aspm(pdev, PCIE_LINK_STATE_L1);

Are all of them safe for the unlocked version of ASPM disable?
Yinghai Lu - May 12, 2011, 11:58 p.m.
On 05/12/2011 04:32 PM, Jesse Barnes wrote:
> On Wed, 11 May 2011 13:23:21 -0700
> Yinghai Lu <yinghai@kernel.org> wrote:
> 
>> On 05/09/2011 02:35 PM, Jesse Barnes wrote:
>>> On Sun, 08 May 2011 11:54:32 -0700
>>> Yinghai Lu <yinghai@kernel.org> wrote:
>>>
>>>>
>>>> Need to use it in _e1000e_disable_aspm.
>>>> when aer happens,
>>>> pci_walk_bus already have down_read(&pci_bus_sem)...
>>>> then report_slot_reset
>>>>         ==> e1000_io_slot_reset
>>>>                 ==> e1000e_disable_aspm
>>>>                         ==> pci_disable_link_state...
>>>>
>>>> We can not use pci_disable_link_state, and it will try to hold pci_bus_sem again.
>>>>
>>>> Try to have __pci_disable_link_state that will not need to hold pci_bus_sem.
>>>
>>> What about the other callers of e1000e_disable_aspm?  Do they already
>>> have the lock held or is it just reset that needs the already locked
>>> version?
>>
>> yes. 
>>
>> there is another version when aspm is not defined. and it does not use any lock. 
>>
>> #ifdef CONFIG_PCIEASPM
>> static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
>> {
>>         pci_disable_link_state(pdev, state);
>> }
>> #else
>> static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
>> {
>>         int pos;
>>         u16 reg16;
>>
>>         /*
>>          * Both device and parent should have the same ASPM setting.
>>          * Disable ASPM in downstream component first and then upstream.
>>          */
>>         pos = pci_pcie_cap(pdev);
>>         pci_read_config_word(pdev, pos + PCI_EXP_LNKCTL, &reg16);
>>         reg16 &= ~state;
>>         pci_write_config_word(pdev, pos + PCI_EXP_LNKCTL, reg16);
>>
>>         if (!pdev->bus->self)
>>                 return;
>>
>>         pos = pci_pcie_cap(pdev->bus->self);
>>         pci_read_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, &reg16);
>>         reg16 &= ~state;
>>         pci_write_config_word(pdev->bus->self, pos + PCI_EXP_LNKCTL, reg16);
>> }
>> #endif
> 
> No, I mean __e1000e_disable_aspm is called from several spots:
> 
> *** drivers/net/e1000e/82571.c:
> e1000_get_variants_82571[435]  e1000e_disable_aspm(adapter->pdev, PCIE_LINK_STATE_L0S);
> 
> *** drivers/net/e1000e/netdev.c:
> e1000_change_mtu[5027]         e1000e_disable_aspm(adapter->pdev, PCIE_LINK_STATE_L1);
> __e1000_resume[5402]           e1000e_disable_aspm(pdev, PCIE_LINK_STATE_L1);
> e1000_io_slot_reset[5650]      e1000e_disable_aspm(pdev, PCIE_LINK_STATE_L1);
> e1000_probe[5797]              e1000e_disable_aspm(pdev, PCIE_LINK_STATE_L1);
> 
> Are all of them safe for the unlocked version of ASPM disable?

yes, there are two version __e1000e_disable_aspm(), one is when aspm support is compiled in, and another one is not.

the one without aspm compiled does not use pci_bus_sem in it self...

So I assume another path should not use pci_bus_sem in the function itself.

Yinghai
--
To unsubscribe from this list: send the line "unsubscribe netdev" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Patch

Index: linux-2.6/drivers/net/e1000e/netdev.c
===================================================================
--- linux-2.6.orig/drivers/net/e1000e/netdev.c
+++ linux-2.6/drivers/net/e1000e/netdev.c
@@ -5360,7 +5360,7 @@  static void e1000_complete_shutdown(stru
 #ifdef CONFIG_PCIEASPM
 static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
 {
-	pci_disable_link_state(pdev, state);
+	__pci_disable_link_state(pdev, state);
 }
 #else
 static void __e1000e_disable_aspm(struct pci_dev *pdev, u16 state)
Index: linux-2.6/drivers/pci/pcie/aspm.c
===================================================================
--- linux-2.6.orig/drivers/pci/pcie/aspm.c
+++ linux-2.6/drivers/pci/pcie/aspm.c
@@ -734,7 +734,7 @@  void pcie_aspm_powersave_config_link(str
  * pci_disable_link_state - disable pci device's link state, so the link will
  * never enter specific states
  */
-void pci_disable_link_state(struct pci_dev *pdev, int state)
+static void ___pci_disable_link_state(struct pci_dev *pdev, int state, bool sem)
 {
 	struct pci_dev *parent = pdev->bus->self;
 	struct pcie_link_state *link;
@@ -747,7 +747,8 @@  void pci_disable_link_state(struct pci_d
 	if (!parent || !parent->link_state)
 		return;
 
-	down_read(&pci_bus_sem);
+	if (sem)
+		down_read(&pci_bus_sem);
 	mutex_lock(&aspm_lock);
 	link = parent->link_state;
 	if (state & PCIE_LINK_STATE_L0S)
@@ -761,7 +762,16 @@  void pci_disable_link_state(struct pci_d
 		pcie_set_clkpm(link, 0);
 	}
 	mutex_unlock(&aspm_lock);
-	up_read(&pci_bus_sem);
+	if (sem)
+		up_read(&pci_bus_sem);
+}
+void __pci_disable_link_state(struct pci_dev *pdev, int state)
+{
+	___pci_disable_link_state(pdev, state, false);
+}
+void pci_disable_link_state(struct pci_dev *pdev, int state)
+{
+	___pci_disable_link_state(pdev, state, true);
 }
 EXPORT_SYMBOL(pci_disable_link_state);
 
Index: linux-2.6/include/linux/pci-aspm.h
===================================================================
--- linux-2.6.orig/include/linux/pci-aspm.h
+++ linux-2.6/include/linux/pci-aspm.h
@@ -28,6 +28,7 @@  extern void pcie_aspm_exit_link_state(st
 extern void pcie_aspm_pm_state_change(struct pci_dev *pdev);
 extern void pcie_aspm_powersave_config_link(struct pci_dev *pdev);
 extern void pci_disable_link_state(struct pci_dev *pdev, int state);
+extern void __pci_disable_link_state(struct pci_dev *pdev, int state);
 extern void pcie_clear_aspm(void);
 extern void pcie_no_aspm(void);
 #else