Re: [PATCH] Revert "PCI: Avoid race while enabling upstream bridges"

From: Konstantin Khlebnikov
Date: Fri Sep 15 2017 - 09:29:02 EST


On 15.09.2017 10:23, Bjorn Helgaas wrote:
This reverts commit 40f11adc7cd9281227f0a6a627d966dd0a5f0cd9.

Jens found that iwlwifi firmware loading failed on a Lenovo X1 Carbon,
gen4:

iwlwifi 0000:04:00.0: Direct firmware load for iwlwifi-8000C-34.ucode failed with error -2
iwlwifi 0000:04:00.0: Direct firmware load for iwlwifi-8000C-33.ucode failed with error -2
iwlwifi 0000:04:00.0: Direct firmware load for iwlwifi-8000C-32.ucode failed with error -2
iwlwifi 0000:04:00.0: loaded firmware version 31.532993.0 op_mode iwlmvm
iwlwifi 0000:04:00.0: Detected Intel(R) Dual Band Wireless AC 8260, REV=0x208
...
iwlwifi 0000:04:00.0: Failed to load firmware chunk!
iwlwifi 0000:04:00.0: Could not load the [0] uCode section
iwlwifi 0000:04:00.0: Failed to start INIT ucode: -110
iwlwifi 0000:04:00.0: Failed to run INIT ucode: -110

He bisected it to 40f11adc7cd9 ("PCI: Avoid race while enabling upstream
bridges"). Revert that commit to fix the regression.

I've found this too. Bug in fast-path: reverting last hunk is enough
150547971091.977464.16294045866179907260.stgit@buzz">http://lkml.kernel.org/r/150547971091.977464.16294045866179907260.stgit@buzz


Link: http://lkml.kernel.org/r/4bcbcbc1-7c79-09f0-5071-bc2f53bf6574@xxxxxxxxx
Fixes: 40f11adc7cd9 ("PCI: Avoid race while enabling upstream bridges")
Signed-off-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx>
CC: Srinath Mannam <srinath.mannam@xxxxxxxxxxxx>
CC: Jens Axboe <axboe@xxxxxxxxx>
CC: Luca Coelho <luca@xxxxxxxxx>
CC: Johannes Berg <johannes@xxxxxxxxxxxxxxxx>
CC: Emmanuel Grumbach <emmanuel.grumbach@xxxxxxxxx>
---
drivers/pci/pci.c | 13 ++-----------
1 file changed, 2 insertions(+), 11 deletions(-)

diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
index b0002daa50f3..6078dfc11b11 100644
--- a/drivers/pci/pci.c
+++ b/drivers/pci/pci.c
@@ -52,7 +52,6 @@ static void pci_pme_list_scan(struct work_struct *work);
static LIST_HEAD(pci_pme_list);
static DEFINE_MUTEX(pci_pme_list_mutex);
static DECLARE_DELAYED_WORK(pci_pme_work, pci_pme_list_scan);
-static DEFINE_MUTEX(pci_bridge_mutex);
struct pci_pme_device {
struct list_head list;
@@ -1351,16 +1350,10 @@ static void pci_enable_bridge(struct pci_dev *dev)
if (bridge)
pci_enable_bridge(bridge);
- /*
- * Hold pci_bridge_mutex to prevent a race when enabling two
- * devices below the bridge simultaneously. The race may cause a
- * PCI_COMMAND_MEMORY update to be lost (see changelog).
- */
- mutex_lock(&pci_bridge_mutex);
if (pci_is_enabled(dev)) {
if (!dev->is_busmaster)
pci_set_master(dev);
- goto end;
+ return;
}
retval = pci_enable_device(dev);
@@ -1368,8 +1361,6 @@ static void pci_enable_bridge(struct pci_dev *dev)
dev_err(&dev->dev, "Error enabling bridge (%d), continuing\n",
retval);
pci_set_master(dev);
-end:
- mutex_unlock(&pci_bridge_mutex);
}
static int pci_enable_device_flags(struct pci_dev *dev, unsigned long flags)
@@ -1394,7 +1385,7 @@ static int pci_enable_device_flags(struct pci_dev *dev, unsigned long flags)
return 0; /* already enabled */
bridge = pci_upstream_bridge(dev);
- if (bridge && !pci_is_enabled(bridge))
+ if (bridge)
pci_enable_bridge(bridge);
/* only skip sriov related */