April 2018 - Linux-stable-mirror

Patch "i40iw: Fix sequence number for the first partial FPDU" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled i40iw: Fix sequence number for the first partial FPDU to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: i40iw-fix-sequence-number-for-the-first-partial-fpdu.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Shiraz Saleem <shiraz.saleem(a)intel.com> Date: Fri, 22 Dec 2017 09:46:59 -0600 Subject: i40iw: Fix sequence number for the first partial FPDU From: Shiraz Saleem <shiraz.saleem(a)intel.com> [ Upstream commit df8b13a1b23356d01dfc4647a5629cdb0f4ce566 ] Partial FPDU processing is broken as the sequence number for the first partial FPDU is wrong due to incorrect Q2 buffer offset. The offset should be 64 rather than 16. Fixes: 786c6adb3a94 ("i40iw: add puda code") Signed-off-by: Shiraz Saleem <shiraz.saleem(a)intel.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/i40iw/i40iw_d.h | 1 + drivers/infiniband/hw/i40iw/i40iw_puda.c | 2 +- 2 files changed, 2 insertions(+), 1 deletion(-) --- a/drivers/infiniband/hw/i40iw/i40iw_d.h +++ b/drivers/infiniband/hw/i40iw/i40iw_d.h @@ -93,6 +93,7 @@ #define RDMA_OPCODE_MASK 0x0f #define RDMA_READ_REQ_OPCODE 1 #define Q2_BAD_FRAME_OFFSET 72 +#define Q2_FPSN_OFFSET 64 #define CQE_MAJOR_DRV 0x8000 #define I40IW_TERM_SENT 0x01 --- a/drivers/infiniband/hw/i40iw/i40iw_puda.c +++ b/drivers/infiniband/hw/i40iw/i40iw_puda.c @@ -1376,7 +1376,7 @@ static void i40iw_ieq_handle_exception(s u32 *hw_host_ctx = (u32 *)qp->hw_host_ctx; u32 rcv_wnd = hw_host_ctx[23]; /* first partial seq # in q2 */ - u32 fps = qp->q2_buf[16]; + u32 fps = *(u32 *)(qp->q2_buf + Q2_FPSN_OFFSET); struct list_head *rxlist = &pfpdu->rxlist; struct list_head *plist; Patches currently in stable-queue which might be from shiraz.saleem(a)intel.com are queue-4.14/i40iw-fix-sequence-number-for-the-first-partial-fpdu.patch queue-4.14/i40iw-correct-q1-xf-object-count-equation.patch queue-4.14/i40iw-validate-correct-ird-ord-connection-parameters.patch

6 years, 9 months

1
0
0 0

Patch "i40iw: Validate correct IRD/ORD connection parameters" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled i40iw: Validate correct IRD/ORD connection parameters to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: i40iw-validate-correct-ird-ord-connection-parameters.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Tatyana Nikolova <tatyana.e.nikolova(a)intel.com> Date: Fri, 22 Dec 2017 09:47:01 -0600 Subject: i40iw: Validate correct IRD/ORD connection parameters From: Tatyana Nikolova <tatyana.e.nikolova(a)intel.com> [ Upstream commit ce9ce74145aa6814a370a9ff4f5a1d719baaced1 ] Casting to u16 before validating IRD/ORD connection parameters could cause recording wrong IRD/ORD values in the cm_node. Validate the IRD/ORD parameters as they are passed by the application before recording them. Fixes: f27b4746f378 ("i40iw: add connection management code") Signed-off-by: Tatyana Nikolova <tatyana.e.nikolova(a)intel.com> Signed-off-by: Shiraz Saleem <shiraz.saleem(a)intel.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/i40iw/i40iw_cm.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) --- a/drivers/infiniband/hw/i40iw/i40iw_cm.c +++ b/drivers/infiniband/hw/i40iw/i40iw_cm.c @@ -125,7 +125,8 @@ static u8 i40iw_derive_hw_ird_setting(u1 * @conn_ird: connection IRD * @conn_ord: connection ORD */ -static void i40iw_record_ird_ord(struct i40iw_cm_node *cm_node, u16 conn_ird, u16 conn_ord) +static void i40iw_record_ird_ord(struct i40iw_cm_node *cm_node, u32 conn_ird, + u32 conn_ord) { if (conn_ird > I40IW_MAX_IRD_SIZE) conn_ird = I40IW_MAX_IRD_SIZE; @@ -3841,7 +3842,7 @@ int i40iw_connect(struct iw_cm_id *cm_id } cm_node->apbvt_set = true; - i40iw_record_ird_ord(cm_node, (u16)conn_param->ird, (u16)conn_param->ord); + i40iw_record_ird_ord(cm_node, conn_param->ird, conn_param->ord); if (cm_node->send_rdma0_op == SEND_RDMA_READ_ZERO && !cm_node->ord_size) cm_node->ord_size = 1; Patches currently in stable-queue which might be from tatyana.e.nikolova(a)intel.com are queue-4.14/i40iw-validate-correct-ird-ord-connection-parameters.patch

6 years, 9 months

1
0
0 0

Patch "i40iw: Correct Q1/XF object count equation" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled i40iw: Correct Q1/XF object count equation to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: i40iw-correct-q1-xf-object-count-equation.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Shiraz Saleem <shiraz.saleem(a)intel.com> Date: Fri, 22 Dec 2017 09:46:56 -0600 Subject: i40iw: Correct Q1/XF object count equation From: Shiraz Saleem <shiraz.saleem(a)intel.com> [ Upstream commit fe99afd1febd74e0ef1fed7e3283f09effe1f4f0 ] Lower Inbound RDMA Read Queue (Q1) object count by a factor of 2 as it is incorrectly doubled. Also, round up Q1 and Transmit FIFO (XF) object count to power of 2 to satisfy hardware requirement. Fixes: 86dbcd0f12e9 ("i40iw: add file to handle cqp calls") Signed-off-by: Shiraz Saleem <shiraz.saleem(a)intel.com> Signed-off-by: Jason Gunthorpe <jgg(a)mellanox.com> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/infiniband/hw/i40iw/i40iw_ctrl.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) --- a/drivers/infiniband/hw/i40iw/i40iw_ctrl.c +++ b/drivers/infiniband/hw/i40iw/i40iw_ctrl.c @@ -3875,8 +3875,10 @@ enum i40iw_status_code i40iw_config_fpm_ hmc_info->hmc_obj[I40IW_HMC_IW_APBVT_ENTRY].cnt = 1; hmc_info->hmc_obj[I40IW_HMC_IW_MR].cnt = mrwanted; - hmc_info->hmc_obj[I40IW_HMC_IW_XF].cnt = I40IW_MAX_WQ_ENTRIES * qpwanted; - hmc_info->hmc_obj[I40IW_HMC_IW_Q1].cnt = 4 * I40IW_MAX_IRD_SIZE * qpwanted; + hmc_info->hmc_obj[I40IW_HMC_IW_XF].cnt = + roundup_pow_of_two(I40IW_MAX_WQ_ENTRIES * qpwanted); + hmc_info->hmc_obj[I40IW_HMC_IW_Q1].cnt = + roundup_pow_of_two(2 * I40IW_MAX_IRD_SIZE * qpwanted); hmc_info->hmc_obj[I40IW_HMC_IW_XFFL].cnt = hmc_info->hmc_obj[I40IW_HMC_IW_XF].cnt / hmc_fpm_misc->xf_block_size; hmc_info->hmc_obj[I40IW_HMC_IW_Q1FL].cnt = Patches currently in stable-queue which might be from shiraz.saleem(a)intel.com are queue-4.14/i40iw-fix-sequence-number-for-the-first-partial-fpdu.patch queue-4.14/i40iw-correct-q1-xf-object-count-equation.patch queue-4.14/i40iw-validate-correct-ird-ord-connection-parameters.patch

6 years, 9 months

1
0
0 0

Patch "i40evf: don't rely on netif_running() outside rtnl_lock()" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled i40evf: don't rely on netif_running() outside rtnl_lock() to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: i40evf-don-t-rely-on-netif_running-outside-rtnl_lock.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Jacob Keller <jacob.e.keller(a)intel.com> Date: Fri, 27 Oct 2017 11:06:49 -0400 Subject: i40evf: don't rely on netif_running() outside rtnl_lock() From: Jacob Keller <jacob.e.keller(a)intel.com> [ Upstream commit 44b034b406211fc103159f82b9e601e05675c739 ] In i40evf_reset_task we use netif_running() to determine whether or not the device is currently up. This allows us to properly free queue memory and shut down things before we request the hardware reset. It turns out that we cannot be guaranteed of netif_running() returning false until the device is fully up, as the kernel core code sets __LINK_STATE_START prior to calling .ndo_open. Since we're not holding the rtnl_lock(), it's possible that the driver's i40evf_open handler function is currently being called while we're resetting. We can't simply hold the rtnl_lock() while checking netif_running() as this could cause a deadlock with the i40evf_open() function. Additionally, we can't avoid the deadlock by holding the rtnl_lock() over the whole reset path, as this essentially serializes all resets, and can cause massive delays if we have multiple VFs on a system. Instead, lets just check our own internal state __I40EVF_RUNNING state field. This allows us to ensure that the state is correct and is only set after we've finished bringing the device up. Without this change we might free data structures about device queues and other memory before they've been fully allocated. Signed-off-by: Jacob Keller <jacob.e.keller(a)intel.com> Tested-by: Andrew Bowers <andrewx.bowers(a)intel.com> Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher(a)intel.com> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/intel/i40evf/i40evf_main.c | 20 +++++++++++++++++--- 1 file changed, 17 insertions(+), 3 deletions(-) --- a/drivers/net/ethernet/intel/i40evf/i40evf_main.c +++ b/drivers/net/ethernet/intel/i40evf/i40evf_main.c @@ -1775,7 +1775,11 @@ static void i40evf_disable_vf(struct i40 adapter->flags |= I40EVF_FLAG_PF_COMMS_FAILED; - if (netif_running(adapter->netdev)) { + /* We don't use netif_running() because it may be true prior to + * ndo_open() returning, so we can't assume it means all our open + * tasks have finished, since we're not holding the rtnl_lock here. + */ + if (adapter->state == __I40EVF_RUNNING) { set_bit(__I40E_VSI_DOWN, adapter->vsi.state); netif_carrier_off(adapter->netdev); netif_tx_disable(adapter->netdev); @@ -1833,6 +1837,7 @@ static void i40evf_reset_task(struct wor struct i40evf_mac_filter *f; u32 reg_val; int i = 0, err; + bool running; while (test_and_set_bit(__I40EVF_IN_CLIENT_TASK, &adapter->crit_section)) @@ -1892,7 +1897,13 @@ static void i40evf_reset_task(struct wor } continue_reset: - if (netif_running(netdev)) { + /* We don't use netif_running() because it may be true prior to + * ndo_open() returning, so we can't assume it means all our open + * tasks have finished, since we're not holding the rtnl_lock here. + */ + running = (adapter->state == __I40EVF_RUNNING); + + if (running) { netif_carrier_off(netdev); netif_tx_stop_all_queues(netdev); adapter->link_up = false; @@ -1936,7 +1947,10 @@ continue_reset: mod_timer(&adapter->watchdog_timer, jiffies + 2); - if (netif_running(adapter->netdev)) { + /* We were running when the reset started, so we need to restore some + * state here. + */ + if (running) { /* allocate transmit descriptors */ err = i40evf_setup_all_tx_resources(adapter); if (err) Patches currently in stable-queue which might be from jacob.e.keller(a)intel.com are queue-4.14/i40evf-don-t-rely-on-netif_running-outside-rtnl_lock.patch

6 years, 9 months

1
0
0 0

Patch "hwmon: (ina2xx) Make calibration register value fixed" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled hwmon: (ina2xx) Make calibration register value fixed to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: hwmon-ina2xx-make-calibration-register-value-fixed.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Maciej Purski <m.purski(a)samsung.com> Date: Wed, 22 Nov 2017 16:32:15 +0100 Subject: hwmon: (ina2xx) Make calibration register value fixed From: Maciej Purski <m.purski(a)samsung.com> [ Upstream commit 5d389b125186cf254ad5b8015763ac07c151aea4 ] Calibration register is used for calculating current register in hardware according to datasheet: current = shunt_volt * calib_register / 2048 (ina 226) current = shunt_volt * calib_register / 4096 (ina 219) Fix calib_register value to 2048 for ina226 and 4096 for ina 219 in order to avoid truncation error and provide best precision allowed by shunt_voltage measurement. Make current scale value follow changes of shunt_resistor from sysfs as calib_register value is now fixed. Power_lsb value should also follow shunt_resistor changes as stated in datasheet: power_lsb = 25 * current_lsb (ina 226) power_lsb = 20 * current_lsb (ina 219) Signed-off-by: Maciej Purski <m.purski(a)samsung.com> Signed-off-by: Guenter Roeck <linux(a)roeck-us.net> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/hwmon/ina2xx.c | 87 ++++++++++++++++++++++++++++--------------------- 1 file changed, 50 insertions(+), 37 deletions(-) --- a/drivers/hwmon/ina2xx.c +++ b/drivers/hwmon/ina2xx.c @@ -95,18 +95,20 @@ enum ina2xx_ids { ina219, ina226 }; struct ina2xx_config { u16 config_default; - int calibration_factor; + int calibration_value; int registers; int shunt_div; int bus_voltage_shift; int bus_voltage_lsb; /* uV */ - int power_lsb; /* uW */ + int power_lsb_factor; }; struct ina2xx_data { const struct ina2xx_config *config; long rshunt; + long current_lsb_uA; + long power_lsb_uW; struct mutex config_lock; struct regmap *regmap; @@ -116,21 +118,21 @@ struct ina2xx_data { static const struct ina2xx_config ina2xx_config[] = { [ina219] = { .config_default = INA219_CONFIG_DEFAULT, - .calibration_factor = 40960000, + .calibration_value = 4096, .registers = INA219_REGISTERS, .shunt_div = 100, .bus_voltage_shift = 3, .bus_voltage_lsb = 4000, - .power_lsb = 20000, + .power_lsb_factor = 20, }, [ina226] = { .config_default = INA226_CONFIG_DEFAULT, - .calibration_factor = 5120000, + .calibration_value = 2048, .registers = INA226_REGISTERS, .shunt_div = 400, .bus_voltage_shift = 0, .bus_voltage_lsb = 1250, - .power_lsb = 25000, + .power_lsb_factor = 25, }, }; @@ -169,12 +171,16 @@ static u16 ina226_interval_to_reg(int in return INA226_SHIFT_AVG(avg_bits); } +/* + * Calibration register is set to the best value, which eliminates + * truncation errors on calculating current register in hardware. + * According to datasheet (eq. 3) the best values are 2048 for + * ina226 and 4096 for ina219. They are hardcoded as calibration_value. + */ static int ina2xx_calibrate(struct ina2xx_data *data) { - u16 val = DIV_ROUND_CLOSEST(data->config->calibration_factor, - data->rshunt); - - return regmap_write(data->regmap, INA2XX_CALIBRATION, val); + return regmap_write(data->regmap, INA2XX_CALIBRATION, + data->config->calibration_value); } /* @@ -187,10 +193,6 @@ static int ina2xx_init(struct ina2xx_dat if (ret < 0) return ret; - /* - * Set current LSB to 1mA, shunt is in uOhms - * (equation 13 in datasheet). - */ return ina2xx_calibrate(data); } @@ -268,15 +270,15 @@ static int ina2xx_get_value(struct ina2x val = DIV_ROUND_CLOSEST(val, 1000); break; case INA2XX_POWER: - val = regval * data->config->power_lsb; + val = regval * data->power_lsb_uW; break; case INA2XX_CURRENT: - /* signed register, LSB=1mA (selected), in mA */ - val = (s16)regval; + /* signed register, result in mA */ + val = regval * data->current_lsb_uA; + val = DIV_ROUND_CLOSEST(val, 1000); break; case INA2XX_CALIBRATION: - val = DIV_ROUND_CLOSEST(data->config->calibration_factor, - regval); + val = regval; break; default: /* programmer goofed */ @@ -304,9 +306,32 @@ static ssize_t ina2xx_show_value(struct ina2xx_get_value(data, attr->index, regval)); } -static ssize_t ina2xx_set_shunt(struct device *dev, - struct device_attribute *da, - const char *buf, size_t count) +/* + * In order to keep calibration register value fixed, the product + * of current_lsb and shunt_resistor should also be fixed and equal + * to shunt_voltage_lsb = 1 / shunt_div multiplied by 10^9 in order + * to keep the scale. + */ +static int ina2xx_set_shunt(struct ina2xx_data *data, long val) +{ + unsigned int dividend = DIV_ROUND_CLOSEST(1000000000, + data->config->shunt_div); + if (val <= 0 || val > dividend) + return -EINVAL; + + mutex_lock(&data->config_lock); + data->rshunt = val; + data->current_lsb_uA = DIV_ROUND_CLOSEST(dividend, val); + data->power_lsb_uW = data->config->power_lsb_factor * + data->current_lsb_uA; + mutex_unlock(&data->config_lock); + + return 0; +} + +static ssize_t ina2xx_store_shunt(struct device *dev, + struct device_attribute *da, + const char *buf, size_t count) { unsigned long val; int status; @@ -316,18 +341,9 @@ static ssize_t ina2xx_set_shunt(struct d if (status < 0) return status; - if (val == 0 || - /* Values greater than the calibration factor make no sense. */ - val > data->config->calibration_factor) - return -EINVAL; - - mutex_lock(&data->config_lock); - data->rshunt = val; - status = ina2xx_calibrate(data); - mutex_unlock(&data->config_lock); + status = ina2xx_set_shunt(data, val); if (status < 0) return status; - return count; } @@ -387,7 +403,7 @@ static SENSOR_DEVICE_ATTR(power1_input, /* shunt resistance */ static SENSOR_DEVICE_ATTR(shunt_resistor, S_IRUGO | S_IWUSR, - ina2xx_show_value, ina2xx_set_shunt, + ina2xx_show_value, ina2xx_store_shunt, INA2XX_CALIBRATION); /* update interval (ina226 only) */ @@ -448,10 +464,7 @@ static int ina2xx_probe(struct i2c_clien val = INA2XX_RSHUNT_DEFAULT; } - if (val <= 0 || val > data->config->calibration_factor) - return -ENODEV; - - data->rshunt = val; + ina2xx_set_shunt(data, val); ina2xx_regmap_config.max_register = data->config->registers; Patches currently in stable-queue which might be from m.purski(a)samsung.com are queue-4.14/hwmon-ina2xx-make-calibration-register-value-fixed.patch

6 years, 9 months

1
0
0 0

Patch "gpio: label descriptors using the device name" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled gpio: label descriptors using the device name to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: gpio-label-descriptors-using-the-device-name.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Linus Walleij <linus.walleij(a)linaro.org> Date: Thu, 4 Jan 2018 22:31:11 +0100 Subject: gpio: label descriptors using the device name From: Linus Walleij <linus.walleij(a)linaro.org> [ Upstream commit 24e78079bf2250874e33da2e7cfbb6db72d3caf4 ] Some GPIO lines appear named "?" in the lsgpio dump due to their requesting drivers not passing a reasonable label. Most typically this happens if a device tree node just defines gpios = <...> and not foo-gpios = <...>, the former gets named "foo" and the latter gets named "?". However the struct device passed in is always valid so let's just label the GPIO with dev_name() on the device if no proper label was passed. Cc: Reported-by: Jason Kridner <jkridner(a)beagleboard.org> Reported-by: Jason Kridner <jkridner(a)beagleboard.org> Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpio/gpiolib.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/drivers/gpio/gpiolib.c +++ b/drivers/gpio/gpiolib.c @@ -3340,7 +3340,8 @@ struct gpio_desc *__must_check gpiod_get return desc; } - status = gpiod_request(desc, con_id); + /* If a connection label was passed use that, else use the device name as label */ + status = gpiod_request(desc, con_id ? con_id : dev_name(dev)); if (status < 0) return ERR_PTR(status); Patches currently in stable-queue which might be from linus.walleij(a)linaro.org are queue-4.14/gpio-label-descriptors-using-the-device-name.patch queue-4.14/gpio-thunderx-fix-error-return-code-in-thunderx_gpio_probe.patch queue-4.14/pinctrl-baytrail-enable-glitch-filter-for-gpios-used-as-interrupts.patch

6 years, 9 months

1
0
0 0

Patch "gpio: thunderx: fix error return code in thunderx_gpio_probe()" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled gpio: thunderx: fix error return code in thunderx_gpio_probe() to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: gpio-thunderx-fix-error-return-code-in-thunderx_gpio_probe.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Wei Yongjun <weiyongjun1(a)huawei.com> Date: Wed, 10 Jan 2018 14:37:13 +0000 Subject: gpio: thunderx: fix error return code in thunderx_gpio_probe() From: Wei Yongjun <weiyongjun1(a)huawei.com> [ Upstream commit 76e28f5ffed82b1e81a86c4eb8d0420515765620 ] Fix to return error code -ENOMEM from the error handling case instead of 0, as done elsewhere in this function. Fixes: 5a2a30024d8c ("gpio: Add gpio driver support for ThunderX and OCTEON-TX") Signed-off-by: Wei Yongjun <weiyongjun1(a)huawei.com> Acked-by: David Daney <david.daney(a)cavium.com> Signed-off-by: Linus Walleij <linus.walleij(a)linaro.org> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/gpio/gpio-thunderx.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) --- a/drivers/gpio/gpio-thunderx.c +++ b/drivers/gpio/gpio-thunderx.c @@ -566,8 +566,10 @@ static int thunderx_gpio_probe(struct pc txgpio->irqd = irq_domain_create_hierarchy(irq_get_irq_data(txgpio->msix_entries[0].vector)->domain, 0, 0, of_node_to_fwnode(dev->of_node), &thunderx_gpio_irqd_ops, txgpio); - if (!txgpio->irqd) + if (!txgpio->irqd) { + err = -ENOMEM; goto out; + } /* Push on irq_data and the domain for each line. */ for (i = 0; i < ngpio; i++) { Patches currently in stable-queue which might be from weiyongjun1(a)huawei.com are queue-4.14/gpio-thunderx-fix-error-return-code-in-thunderx_gpio_probe.patch

6 years, 9 months

1
0
0 0

Patch "cxgb4vf: Fix SGE FL buffer initialization logic for 64K pages" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled cxgb4vf: Fix SGE FL buffer initialization logic for 64K pages to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: cxgb4vf-fix-sge-fl-buffer-initialization-logic-for-64k-pages.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Arjun Vynipadath <arjun(a)chelsio.com> Date: Wed, 10 Jan 2018 12:02:13 +0530 Subject: cxgb4vf: Fix SGE FL buffer initialization logic for 64K pages From: Arjun Vynipadath <arjun(a)chelsio.com> [ Upstream commit ea0a42109aee7b92e631c4eb3f2219fadf58acdd ] We'd come in with SGE_FL_BUFFER_SIZE[0] and [1] both equal to 64KB and the extant logic would flag that as an error. This was already fixed in cxgb4 driver with "92ddcc7 cxgb4: Fix some small bugs in t4_sge_init_soft() when our Page Size is 64KB". Original Work by: Casey Leedom <leedom(a)chelsio.com> Signed-off-by: Arjun Vynipadath <arjun(a)chelsio.com> Signed-off-by: Ganesh Goudar <ganeshgr(a)chelsio.com> Signed-off-by: David S. Miller <davem(a)davemloft.net> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/net/ethernet/chelsio/cxgb4vf/sge.c | 23 +++++++++++++++++------ 1 file changed, 17 insertions(+), 6 deletions(-) --- a/drivers/net/ethernet/chelsio/cxgb4vf/sge.c +++ b/drivers/net/ethernet/chelsio/cxgb4vf/sge.c @@ -2619,8 +2619,8 @@ void t4vf_sge_stop(struct adapter *adapt int t4vf_sge_init(struct adapter *adapter) { struct sge_params *sge_params = &adapter->params.sge; - u32 fl0 = sge_params->sge_fl_buffer_size[0]; - u32 fl1 = sge_params->sge_fl_buffer_size[1]; + u32 fl_small_pg = sge_params->sge_fl_buffer_size[0]; + u32 fl_large_pg = sge_params->sge_fl_buffer_size[1]; struct sge *s = &adapter->sge; /* @@ -2628,9 +2628,20 @@ int t4vf_sge_init(struct adapter *adapte * the Physical Function Driver. Ideally we should be able to deal * with _any_ configuration. Practice is different ... */ - if (fl0 != PAGE_SIZE || (fl1 != 0 && fl1 <= fl0)) { + + /* We only bother using the Large Page logic if the Large Page Buffer + * is larger than our Page Size Buffer. + */ + if (fl_large_pg <= fl_small_pg) + fl_large_pg = 0; + + /* The Page Size Buffer must be exactly equal to our Page Size and the + * Large Page Size Buffer should be 0 (per above) or a power of 2. + */ + if (fl_small_pg != PAGE_SIZE || + (fl_large_pg & (fl_large_pg - 1)) != 0) { dev_err(adapter->pdev_dev, "bad SGE FL buffer sizes [%d, %d]\n", - fl0, fl1); + fl_small_pg, fl_large_pg); return -EINVAL; } if ((sge_params->sge_control & RXPKTCPLMODE_F) != @@ -2642,8 +2653,8 @@ int t4vf_sge_init(struct adapter *adapte /* * Now translate the adapter parameters into our internal forms. */ - if (fl1) - s->fl_pg_order = ilog2(fl1) - PAGE_SHIFT; + if (fl_large_pg) + s->fl_pg_order = ilog2(fl_large_pg) - PAGE_SHIFT; s->stat_len = ((sge_params->sge_control & EGRSTATUSPAGESIZE_F) ? 128 : 64); s->pktshift = PKTSHIFT_G(sge_params->sge_control); Patches currently in stable-queue which might be from arjun(a)chelsio.com are queue-4.14/cxgb4vf-fix-sge-fl-buffer-initialization-logic-for-64k-pages.patch

6 years, 9 months

1
0
0 0

Patch "EDAC, mv64x60: Fix an error handling path" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled EDAC, mv64x60: Fix an error handling path to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: edac-mv64x60-fix-an-error-handling-path.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> Date: Sun, 7 Jan 2018 21:54:00 +0100 Subject: EDAC, mv64x60: Fix an error handling path From: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> [ Upstream commit 68fa24f9121c04ef146b5158f538c8b32f285be5 ] We should not call edac_mc_del_mc() if a corresponding call to edac_mc_add_mc() has not been performed yet. So here, we should go to err instead of err2 to branch at the right place of the error handling path. Signed-off-by: Christophe JAILLET <christophe.jaillet(a)wanadoo.fr> Cc: linux-edac <linux-edac(a)vger.kernel.org> Link: http://lkml.kernel.org/r/20180107205400.14068-1-christophe.jaillet@wanadoo.… Signed-off-by: Borislav Petkov <bp(a)suse.de> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- drivers/edac/mv64x60_edac.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/drivers/edac/mv64x60_edac.c +++ b/drivers/edac/mv64x60_edac.c @@ -758,7 +758,7 @@ static int mv64x60_mc_err_probe(struct p /* Non-ECC RAM? */ printk(KERN_WARNING "%s: No ECC DIMMs discovered\n", __func__); res = -ENODEV; - goto err2; + goto err; } edac_dbg(3, "init mci\n"); Patches currently in stable-queue which might be from christophe.jaillet(a)wanadoo.fr are queue-4.14/edac-mv64x60-fix-an-error-handling-path.patch queue-4.14/asoc-intel-sst-fix-the-return-value-of-sst_send_byte_stream_mrfld.patch

6 years, 9 months

1
0
0 0

Patch "crypto: aes-generic - build with -Os on gcc-7+" has been added to the 4.14-stable tree

by gregkh＠linuxfoundation.org

This is a note to let you know that I've just added the patch titled crypto: aes-generic - build with -Os on gcc-7+ to the 4.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=sum… The filename of the patch is: crypto-aes-generic-build-with-os-on-gcc-7.patch and it can be found in the queue-4.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable(a)vger.kernel.org> know about it. >From foo@baz Mon Apr 9 13:58:16 CEST 2018 From: Arnd Bergmann <arnd(a)arndb.de> Date: Wed, 3 Jan 2018 23:39:27 +0100 Subject: crypto: aes-generic - build with -Os on gcc-7+ From: Arnd Bergmann <arnd(a)arndb.de> [ Upstream commit 148b974deea927f5dbb6c468af2707b488bfa2de ] While testing other changes, I discovered that gcc-7.2.1 produces badly optimized code for aes_encrypt/aes_decrypt. This is especially true when CONFIG_UBSAN_SANITIZE_ALL is enabled, where it leads to extremely large stack usage that in turn might cause kernel stack overflows: crypto/aes_generic.c: In function 'aes_encrypt': crypto/aes_generic.c:1371:1: warning: the frame size of 4880 bytes is larger than 2048 bytes [-Wframe-larger-than=] crypto/aes_generic.c: In function 'aes_decrypt': crypto/aes_generic.c:1441:1: warning: the frame size of 4864 bytes is larger than 2048 bytes [-Wframe-larger-than=] I verified that this problem exists on all architectures that are supported by gcc-7.2, though arm64 in particular is less affected than the others. I also found that gcc-7.1 and gcc-8 do not show the extreme stack usage but still produce worse code than earlier versions for this file, apparently because of optimization passes that generally provide a substantial improvement in object code quality but understandably fail to find any shortcuts in the AES algorithm. Possible workarounds include a) disabling -ftree-pre and -ftree-sra optimizations, this was an earlier patch I tried, which reliably fixed the stack usage, but caused a serious performance regression in some versions, as later testing found. b) disabling UBSAN on this file or all ciphers, as suggested by Ard Biesheuvel. This would lead to massively better crypto performance in UBSAN-enabled kernels and avoid the stack usage, but there is a concern over whether we should exclude arbitrary files from UBSAN at all. c) Forcing the optimization level in a different way. Similar to a), but rather than deselecting specific optimization stages, this now uses "gcc -Os" for this file, regardless of the CONFIG_CC_OPTIMIZE_FOR_PERFORMANCE/SIZE option. This is a reliable workaround for the stack consumption on all architecture, and I've retested the performance results now on x86, cycles/byte (lower is better) for cbc(aes-generic) with 256 bit keys: -O2 -Os gcc-6.3.1 14.9 15.1 gcc-7.0.1 14.7 15.3 gcc-7.1.1 15.3 14.7 gcc-7.2.1 16.8 15.9 gcc-8.0.0 15.5 15.6 This implements the option c) by enabling forcing -Os on all compiler versions starting with gcc-7.1. As a workaround for PR83356, it would only be needed for gcc-7.2+ with UBSAN enabled, but since it also shows better performance on gcc-7.1 without UBSAN, it seems appropriate to use the faster version here as well. Side note: during testing, I also played with the AES code in libressl, which had a similar performance regression from gcc-6 to gcc-7.2, but was three times slower overall. It might be interesting to investigate that further and possibly port the Linux implementation into that. Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83356 Link: https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83651 Cc: Richard Biener <rguenther(a)suse.de> Cc: Jakub Jelinek <jakub(a)gcc.gnu.org> Cc: Ard Biesheuvel <ard.biesheuvel(a)linaro.org> Signed-off-by: Arnd Bergmann <arnd(a)arndb.de> Acked-by: Ard Biesheuvel <ard.biesheuvel(a)linaro.org> Signed-off-by: Herbert Xu <herbert(a)gondor.apana.org.au> Signed-off-by: Sasha Levin <alexander.levin(a)microsoft.com> Signed-off-by: Greg Kroah-Hartman <gregkh(a)linuxfoundation.org> --- crypto/Makefile | 1 + 1 file changed, 1 insertion(+) --- a/crypto/Makefile +++ b/crypto/Makefile @@ -98,6 +98,7 @@ obj-$(CONFIG_CRYPTO_TWOFISH_COMMON) += t obj-$(CONFIG_CRYPTO_SERPENT) += serpent_generic.o CFLAGS_serpent_generic.o := $(call cc-option,-fsched-pressure) # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79149 obj-$(CONFIG_CRYPTO_AES) += aes_generic.o +CFLAGS_aes_generic.o := $(call cc-ifversion, -ge, 0701, -Os) # https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83356 obj-$(CONFIG_CRYPTO_AES_TI) += aes_ti.o obj-$(CONFIG_CRYPTO_CAMELLIA) += camellia_generic.o obj-$(CONFIG_CRYPTO_CAST_COMMON) += cast_common.o Patches currently in stable-queue which might be from arnd(a)arndb.de are queue-4.14/crypto-aes-generic-build-with-os-on-gcc-7.patch

6 years, 9 months

1
0
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror April 2018