From: Shuchang Li <[email protected]>
When if_type equals to zero and pci_resource_start(pdev, PCI_64BIT_BAR4)
returns false, drbl_regs_memmap_p is not remapped. However, in the code
it goes to out_iounmap_all, where drbl_regs_memmap_p is unmapped. This
may cause some problems.
When if_type equals to six and pci_resource_start(pdev, PCI_64BIT_BAR4)
returns true, drbl_regs_memmap_p may has been remapped and
ctrl_regs_memmap_p is not remapped. However in the code it goes to
out_iounmap_ctrl, where ctrl_regs_memmap_p is unmapped and
drbl_regs_memmap_p is not unmapped. This may cause a leak and
some other problems.
To fix these issues, we need to add null checks before iounmap(), and
change some goto lables.
Signed-off-by: Shuchang Li <[email protected]>
Reviewed-by: Dongliang Mu <[email protected]>
---
drivers/scsi/lpfc/lpfc_init.c | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/drivers/scsi/lpfc/lpfc_init.c b/drivers/scsi/lpfc/lpfc_init.c
index 4f7485958c49..bd5e734487e3 100644
--- a/drivers/scsi/lpfc/lpfc_init.c
+++ b/drivers/scsi/lpfc/lpfc_init.c
@@ -12026,7 +12026,7 @@ lpfc_sli4_pci_mem_setup(struct lpfc_hba *phba)
goto out_iounmap_all;
} else {
error = -ENOMEM;
- goto out_iounmap_all;
+ goto out_iounmap_ctrl;
}
}
@@ -12044,7 +12044,7 @@ lpfc_sli4_pci_mem_setup(struct lpfc_hba *phba)
dev_err(&pdev->dev,
"ioremap failed for SLI4 HBA dpp registers.\n");
error = -ENOMEM;
- goto out_iounmap_ctrl;
+ goto out_iounmap_all;
}
phba->pci_bar4_memmap_p = phba->sli4_hba.dpp_regs_memmap_p;
}
@@ -12069,9 +12069,11 @@ lpfc_sli4_pci_mem_setup(struct lpfc_hba *phba)
return 0;
out_iounmap_all:
- iounmap(phba->sli4_hba.drbl_regs_memmap_p);
+ if (!phba->sli4_hba.drbl_regs_memmap_p)
+ iounmap(phba->sli4_hba.drbl_regs_memmap_p);
out_iounmap_ctrl:
- iounmap(phba->sli4_hba.ctrl_regs_memmap_p);
+ if (!phba->sli4_hba.ctrl_regs_memmap_p)
+ iounmap(phba->sli4_hba.ctrl_regs_memmap_p);
out_iounmap_conf:
iounmap(phba->sli4_hba.conf_regs_memmap_p);
--
2.25.1
On Mon, Apr 03, 2023 at 03:48:21PM +0800, [email protected] wrote:
> @@ -12069,9 +12069,11 @@ lpfc_sli4_pci_mem_setup(struct lpfc_hba *phba)
> return 0;
>
> out_iounmap_all:
> - iounmap(phba->sli4_hba.drbl_regs_memmap_p);
> + if (!phba->sli4_hba.drbl_regs_memmap_p)
> + iounmap(phba->sli4_hba.drbl_regs_memmap_p);
The test is reversed still.
If you make a mistake, you should write a static checker warning so that
you never make the same mistake again. ;) See attached.
> out_iounmap_ctrl:
> - iounmap(phba->sli4_hba.ctrl_regs_memmap_p);
> + if (!phba->sli4_hba.ctrl_regs_memmap_p)
Also reversed.
> + iounmap(phba->sli4_hba.ctrl_regs_memmap_p);
> out_iounmap_conf:
> iounmap(phba->sli4_hba.conf_regs_memmap_p);
regards,
dan carpenter
On 2023/4/3 17:16, Dan Carpenter wrote:
> On Mon, Apr 03, 2023 at 03:48:21PM +0800, [email protected] wrote:
>> @@ -12069,9 +12069,11 @@ lpfc_sli4_pci_mem_setup(struct lpfc_hba *phba)
>> return 0;
>>
>> out_iounmap_all:
>> - iounmap(phba->sli4_hba.drbl_regs_memmap_p);
>> + if (!phba->sli4_hba.drbl_regs_memmap_p)
>> + iounmap(phba->sli4_hba.drbl_regs_memmap_p);
> The test is reversed still.
Thanks for your review, Dan. Sorry for my internal careless review.
Shuchang is creating a v2 patch to fix all the mentioned issues.
Really sorry about this stupid mistake.
>
> If you make a mistake, you should write a static checker warning so that
> you never make the same mistake again. ;) See attached.
>
>
>> out_iounmap_ctrl:
>> - iounmap(phba->sli4_hba.ctrl_regs_memmap_p);
>> + if (!phba->sli4_hba.ctrl_regs_memmap_p)
> Also reversed.
>
>> + iounmap(phba->sli4_hba.ctrl_regs_memmap_p);
>> out_iounmap_conf:
>> iounmap(phba->sli4_hba.conf_regs_memmap_p);
> regards,
> dan carpenter
>
>