From: Wang Hai <wanghai38@huawei.com>
Date: Tue, 12 Oct 2021 07:15:56 +0000 (+0800)
Subject: PCI/MSI: Handle msi_populate_sysfs() errors correctly
X-Git-Tag: accepted/tizen/unified/20230118.172025~6183^2
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=2b94b6b79b7c24092a6169db9e83c4565be0db42;p=platform%2Fkernel%2Flinux-rpi.git

PCI/MSI: Handle msi_populate_sysfs() errors correctly

Previously, when msi_populate_sysfs() failed, we saved the error return
value as dev->msi_irq_groups, which leads to a page fault when
free_msi_irqs() calls msi_destroy_sysfs().

To prevent this, leave dev->msi_irq_groups alone when msi_populate_sysfs()
fails.

Found by the Hulk Robot when injecting a memory allocation fault in
msi_populate_sysfs():

  BUG: unable to handle page fault for address: fffffffffffffff4
  ...
  Call Trace:
   msi_destroy_sysfs+0x30/0xa0
   free_msi_irqs+0x11d/0x1b0

Fixes: 2f170814bdd2 ("genirq/msi: Move MSI sysfs handling from PCI to MSI core")
Link: https://lore.kernel.org/r/20211012071556.939137-1-wanghai38@huawei.com
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Wang Hai <wanghai38@huawei.com>
Signed-off-by: Bjorn Helgaas <bhelgaas@google.com>
Acked-by: Barry Song <song.bao.hua@hisilicon.com>
---

diff --git a/drivers/pci/msi.c b/drivers/pci/msi.c
index 0099a00..4b47929 100644
--- a/drivers/pci/msi.c
+++ b/drivers/pci/msi.c
@@ -535,6 +535,7 @@ static int msi_verify_entries(struct pci_dev *dev)
 static int msi_capability_init(struct pci_dev *dev, int nvec,
 			       struct irq_affinity *affd)
 {
+	const struct attribute_group **groups;
 	struct msi_desc *entry;
 	int ret;
 
@@ -558,12 +559,14 @@ static int msi_capability_init(struct pci_dev *dev, int nvec,
 	if (ret)
 		goto err;
 
-	dev->msi_irq_groups = msi_populate_sysfs(&dev->dev);
-	if (IS_ERR(dev->msi_irq_groups)) {
-		ret = PTR_ERR(dev->msi_irq_groups);
+	groups = msi_populate_sysfs(&dev->dev);
+	if (IS_ERR(groups)) {
+		ret = PTR_ERR(groups);
 		goto err;
 	}
 
+	dev->msi_irq_groups = groups;
+
 	/* Set MSI enabled bits	*/
 	pci_intx_for_msi(dev, 0);
 	pci_msi_set_enable(dev, 1);
@@ -691,6 +694,7 @@ static void msix_mask_all(void __iomem *base, int tsize)
 static int msix_capability_init(struct pci_dev *dev, struct msix_entry *entries,
 				int nvec, struct irq_affinity *affd)
 {
+	const struct attribute_group **groups;
 	void __iomem *base;
 	int ret, tsize;
 	u16 control;
@@ -730,12 +734,14 @@ static int msix_capability_init(struct pci_dev *dev, struct msix_entry *entries,
 
 	msix_update_entries(dev, entries);
 
-	dev->msi_irq_groups = msi_populate_sysfs(&dev->dev);
-	if (IS_ERR(dev->msi_irq_groups)) {
-		ret = PTR_ERR(dev->msi_irq_groups);
+	groups = msi_populate_sysfs(&dev->dev);
+	if (IS_ERR(groups)) {
+		ret = PTR_ERR(groups);
 		goto out_free;
 	}
 
+	dev->msi_irq_groups = groups;
+
 	/* Set MSI-X enabled bits and unmask the function */
 	pci_intx_for_msi(dev, 0);
 	dev->msix_enabled = 1;