[BUG] net: broadcom: b44: Possible sleep-in-atomic-context bugs in b44_chip_reset()

[Jia-Ju Bai]

The driver may sleep with holding a spinlock.
The function call paths (from bottom to top) in Linux-4.17 are:

[FUNC] msleep
drivers/ssb/driver_pcicore.c, 671:
    msleep in ssb_pcie_mdio_write
drivers/ssb/driver_pcicore.c, 446:
    ssb_pcie_mdio_write in ssb_pcicore_serdes_workaround
drivers/ssb/driver_pcicore.c, 508:
    ssb_pcicore_serdes_workaround in ssb_pcicore_pcie_setup_workarounds
drivers/ssb/driver_pcicore.c, 729:
    ssb_pcicore_pcie_setup_workarounds in ssb_pcicore_dev_irqvecs_enable
drivers/net/ethernet/broadcom/b44.c, 1289:
    ssb_pcicore_dev_irqvecs_enable in b44_chip_reset
drivers/net/ethernet/broadcom/b44.c, 1363:
    b44_chip_reset in b44_halt
drivers/net/ethernet/broadcom/b44.c, 2498:
    b44_halt in b44_suspend
drivers/net/ethernet/broadcom/b44.c, 2496:
    spin_lock_irq in b44_suspend

[FUNC] msleep
drivers/ssb/driver_pcicore.c, 671:
    msleep in ssb_pcie_mdio_write
drivers/ssb/driver_pcicore.c, 446:
    ssb_pcie_mdio_write in ssb_pcicore_serdes_workaround
drivers/ssb/driver_pcicore.c, 508:
    ssb_pcicore_serdes_workaround in ssb_pcicore_pcie_setup_workarounds
drivers/ssb/driver_pcicore.c, 729:
    ssb_pcicore_pcie_setup_workarounds in ssb_pcicore_dev_irqvecs_enable
drivers/net/ethernet/broadcom/b44.c, 1289:
    ssb_pcicore_dev_irqvecs_enable in b44_chip_reset
drivers/net/ethernet/broadcom/b44.c, 1412:
    b44_chip_reset in b44_init_hw
drivers/net/ethernet/broadcom/b44.c, 2533:
    b44_init_hw in b44_resume
drivers/net/ethernet/broadcom/b44.c, 2531:
    spin_lock_irq in b44_resume

These bugs are found by my static analysis tool DSAC.


Best wishes,
Jia-Ju Bai