3.16.35-rc1 review patch. If anyone has any objections, please let me know. ------------------ From: Lucas Stach <l.stach@xxxxxxxxxxxxxx> commit 4d107d3b5a686b5834e533a00b73bf7b1cf59df7 upstream. imx6_pcie_link_up() previously used usleep_range() to wait for the link to come up. Since it may be called while holding the config spinlock, the sleep causes a "BUG: scheduling while atomic" error. Instead of waiting for the link to come up in imx6_pcie_link_up(), do the waiting in imx6_pcie_wait_for_link(), where we're not holding a lock and sleeping is allowed. [bhelgaas: changelog, references to bugzilla and f95d3ae77191] Link: https://bugzilla.kernel.org/show_bug.cgi?id=100031 Fixes: f95d3ae77191 ("PCI: imx6: Wait for retraining") Signed-off-by: Lucas Stach <l.stach@xxxxxxxxxxxxxx> Signed-off-by: Bjorn Helgaas <bhelgaas@xxxxxxxxxx> [bwh: Backported to 3.16: also update the retry loop in imx6_pcie_wait_for_link() as done upstream in commit 6cbb247e85eb ("PCI: designware: Wait for link to come up with consistent style")] Signed-off-by: Ben Hutchings <ben@xxxxxxxxxxxxxxx> --- drivers/pci/host/pci-imx6.c | 60 +++++++++++++++++---------------------------- 1 file changed, 22 insertions(+), 38 deletions(-) --- a/drivers/pci/host/pci-imx6.c +++ b/drivers/pci/host/pci-imx6.c @@ -298,17 +298,33 @@ static void imx6_pcie_init_phy(struct pc static int imx6_pcie_wait_for_link(struct pcie_port *pp) { - int count = 200; + unsigned int retries; - while (!dw_pcie_link_up(pp)) { - usleep_range(100, 1000); - if (--count) - continue; - - return -EINVAL; + /* + * Test if the PHY reports that the link is up and also that the LTSSM + * training finished. There are three possible states of the link when + * this code is called: + * 1) The link is DOWN (unlikely) + * The link didn't come up yet for some reason. This usually means + * we have a real problem somewhere, if it happens with a peripheral + * connected. This state calls for inspection of the DEBUG registers. + * 2) The link is UP, but still in LTSSM training + * Wait for the training to finish, which should take a very short + * time. If the training does not finish, we have a problem and we + * need to inspect the DEBUG registers. If the training does finish, + * the link is up and operating correctly. + * 3) The link is UP and no longer in LTSSM training + * The link is up and operating correctly. + */ + for (retries = 0; retries < 200; retries++) { + u32 reg = readl(pp->dbi_base + PCIE_PHY_DEBUG_R1); + if ((reg & PCIE_PHY_DEBUG_R1_XMLH_LINK_UP) && + !(reg & PCIE_PHY_DEBUG_R1_XMLH_LINK_IN_TRAINING)) + return 0; + usleep_range(1000, 2000); } - return 0; + return -EINVAL; } static irqreturn_t imx6_pcie_msi_handler(int irq, void *arg) -- To unsubscribe from this list: send the line "unsubscribe stable" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html