New: Function twa_probe() does not set error codes on failures

https://bugzilla.kernel.org/show_bug.cgi?id=188851

Bug ID: 188851
Summary: Function twa_probe() does not set error codes on
failures
Product: SCSI Drivers
Version: 2.5
Kernel Version: linux-4.9-rc6
Hardware: All
OS: Linux
Tree: Mainline
Status: NEW
Severity: normal
Priority: P1
Component: Other
Assignee: [email protected]
Reporter: [email protected]
Regression: No

In function twa_probe(), variable retval takes the error code. However, the error code is not set on some failures. I am not sure whether it is the intention of the author. I list 3 positions that seems anomalous. (1) error code "-ENODEV" is not assigned to retval when the call to twa_initialize_device_extension() (at line 2041) fails; (2) error code "-ENOMEM" is not assigned to retval when the call to ioremap() (at line 2062) fails; (3) error code "-ENODEV" is not assigned to retval when the call to twa_reset_sequence() (at line 2072) fails. Codes related to these bugs are summarised as follows.

twa_probe @@ drivers/scsi/3w-9xxx.c
2003 /* This function will probe and initialize a card */
2004 static int twa_probe(struct pci_dev *pdev, const struct pci_device_id *dev_id)
2005 {

2006     struct Scsi_Host *host = NULL;
2007     TW_Device_Extension *tw_dev;
2008     unsigned long mem_addr, mem_len;
2009     int retval = -ENODEV;

2010
2011 retval = pci_enable_device(pdev);
2012 if (retval) {
2013 TW_PRINTK(host, TW_DRIVER, 0x34, "Failed to enable pci device");
2014 goto out_disable_device;
2015 }

         ...
2041     if (twa_initialize_device_extension(tw_dev)) {
2042         TW_PRINTK(tw_dev->host, TW_DRIVER, 0x25, "Failed to initialize
device extension");
             // (1) The value of retval is 0. Insert "retval = -ENODEV;" here?
2043         goto out_free_device_extension;
2044     }
2045 
2046     /* Request IO regions */
2047     retval = pci_request_regions(pdev, "3w-9xxx");
2048     if (retval) {

2049 TW_PRINTK(tw_dev->host, TW_DRIVER, 0x26, "Failed to get mem
region");
2050 goto out_free_device_extension;
2051 }

         ...
2062     tw_dev->base_addr = ioremap(mem_addr, mem_len);
2063     if (!tw_dev->base_addr) {
2064         TW_PRINTK(tw_dev->host, TW_DRIVER, 0x35, "Failed to ioremap");
             // (2) The value of retval is 0. Insert "retval = -ENOMEM;" here?

2065 goto out_release_mem_region;
2066 }
2067
2068 /* Disable interrupts on the card */
2069 TW_DISABLE_INTERRUPTS(tw_dev);
2070
2071 /* Initialize the card */
2072 if (twa_reset_sequence(tw_dev, 0))

             // (3) The value of retval is 0. Insert "retval = -ENODEV;" here?
2073         goto out_iounmap;
         ...

2133 return 0;
2134
2135 out_remove_host:
2136 if (test_bit(TW_USING_MSI, &tw_dev->flags))
2137 pci_disable_msi(pdev);
2138 scsi_remove_host(host);
2139 out_iounmap:
2140 iounmap(tw_dev->base_addr);
2141 out_release_mem_region:
2142 pci_release_regions(pdev);
2143 out_free_device_extension:
2144 twa_free_device_extension(tw_dev);
2145 scsi_host_put(host);
2146 out_disable_device:
2147 pci_disable_device(pdev);
2148
2149 return retval;
2150 } /* End twa_probe() */

Thanks very much!

You are receiving this mail because:
You are watching the assignee of the bug.