mfd: ti_am335x_tscadc: Use PLATFORM_DEVID_AUTO while registering mfd cells
Subsystems (10)
Sasha Levin
maintainerPATCH 1/34 mfd: ti_am335x_tscadc: Use PLATFORM_DEVID_AUTO while registering mfd cells
From: Vignesh R <[email protected]>
[ Upstream commit b40ee006fe6a8a25093434e5d394128c356a48f3 ]
Use PLATFORM_DEVID_AUTO to number mfd cells while registering, so that different instances are uniquely identified. This is required in order to support registering of multiple instances of same ti_am335x_tscadc IP.
--- drivers/mfd/ti_am335x_tscadc.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
drivers/mfd/ti_am335x_tscadc.c +3 -2
@@ -265,8 +265,9 @@ static int ti_tscadc_probe(struct platform_device *pdev) cell->pdata_size =sizeof (tscadc); } - err =mfd_add_devices (&pdev->dev, pdev->id, tscadc->cells, - tscadc->used_cells,NULL ,0 ,NULL ); + err =mfd_add_devices (&pdev->dev,PLATFORM_DEVID_AUTO , + tscadc->cells, tscadc->used_cells,NULL , +0 ,NULL );if (err <0 ) goto err_disable_clk;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 2/34 pvcalls-back: set -ENOTCONN in pvcalls_conn_back_read
From: Stefano Stabellini <[email protected]>
[ Upstream commit e6587cdbd732eacb4c7ce592ed46f7bbcefb655f ]
When a connection is closing we receive on pvcalls_sk_state_change notification. Instead of setting the connection as closed immediately (-ENOTCONN), let's read one more time from it: pvcalls_conn_back_read will set the connection as closed when necessary.
That way, we avoid races between pvcalls_sk_state_change and pvcalls_back_ioworker.
--- drivers/xen/pvcalls-back.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)
drivers/xen/pvcalls-back.c +4 -3
@@ -164,9 +164,10 @@ static void pvcalls_conn_back_read(void *opaque)/* write the data, then modify the indexes */ virt_wmb (); -if (ret <0 ) +if (ret <0 ) { +atomic_set (&map->read,0 ); intf->in_error = ret; -else + }else intf->in_prod = prod + ret;/* update the indexes, then notify the other end */ virt_wmb ();
@@ -296,7 +297,7 @@ static void pvcalls_sk_state_change(struct sock *sock)return ; intf = map->ring; - intf->in_error = -ENOTCONN ; +atomic_inc (&map->read);notify_remote_via_irq (map->irq); }
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 3/34 mfd: twl-core: Fix section annotations on {,un}protect_pm_master
In reply to Sasha LevinFrom: Nathan Chancellor <[email protected]>
[ Upstream commit 8838555089f0345b87f4277fe5a8dd647dc65589 ]
When building the kernel with Clang, the following section mismatch warning appears:
WARNING: vmlinux.o(.text+0x3d84a3b): Section mismatch in reference from the function twl_probe() to the function .init.text:unprotect_pm_master()
The function twl_probe() references
the function __init unprotect_pm_master().
This is often because twl_probe lacks a __init
annotation or the annotation of unprotect_pm_master is wrong.
Remove the __init annotation on the *protect_pm_master functions so there is no more mismatch.
--- drivers/mfd/twl-core.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/mfd/twl-core.c +2 -2
@@ -979,7 +979,7 @@ add_children(struct twl4030_platform_data *pdata, unsigned irq_base,* letting it generate the right frequencies for USB, MADC, and * other purposes. */ -static inline int __init protect_pm_master (void ) +static inline int protect_pm_master (void ) {int e =0 ;
@@ -988,7 +988,7 @@ static inline int __init protect_pm_master(void)return e; } -static inline int __init unprotect_pm_master (void ) +static inline int unprotect_pm_master (void ) {int e =0 ;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 4/34 mfd: db8500-prcmu: Fix some section annotations
In reply to Sasha LevinFrom: Nathan Chancellor <[email protected]>
[ Upstream commit a3888f62fe66429fad3be7f2ba962e1e08c26fd6 ]
When building the kernel with Clang, the following section mismatch warnings appear:
WARNING: vmlinux.o(.text+0x7239cc): Section mismatch in reference from the function db8500_prcmu_probe() to the function .init.text:init_prcm_registers()
The function db8500_prcmu_probe() references
the function __init init_prcm_registers().
This is often because db8500_prcmu_probe lacks a __init annotation or the annotation of init_prcm_registers is wrong.
WARNING: vmlinux.o(.text+0x723e28): Section mismatch in reference from the function db8500_prcmu_probe() to the function .init.text:fw_project_name()
The function db8500_prcmu_probe() references
the function __init fw_project_name().
This is often because db8500_prcmu_probe lacks a __init annotation or the annotation of fw_project_name is wrong.
db8500_prcmu_probe should not be marked as __init so remove the __init annotation from fw_project_name and init_prcm_registers.
--- drivers/mfd/db8500-prcmu.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/mfd/db8500-prcmu.c +2 -2
@@ -2584,7 +2584,7 @@ static struct irq_chip prcmu_irq_chip = { .irq_unmask = prcmu_irq_unmask, }; -static __init char *fw_project_name (u32 project) +static char *fw_project_name (u32 project) {switch (project) {case PRCMU_FW_PROJECT_U8500 :
@@ -2732,7 +2732,7 @@ void __init db8500_prcmu_early_init(u32 phy_base, u32 size)INIT_WORK (&mb0_transfer.mask_work, prcmu_mask_work); } -static void __init init_prcm_registers (void ) +static void init_prcm_registers (void ) {u32 val;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 5/34 mfd: mt6397: Do not call irq_domain_remove if PMIC unsupported
In reply to Sasha LevinFrom: Nicolas Boichat <[email protected]>
[ Upstream commit a177276aa098aa47a100d51a13eaaef029604b6d ]
If the PMIC ID is unknown, the current code would call irq_domain_remove and panic, as pmic->irq_domain is only initialized by mt6397_irq_init.
Return immediately with an error, if the chip ID is unsupported.
--- drivers/mfd/mt6397-core.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/mfd/mt6397-core.c +1 -2
@@ -309,8 +309,7 @@ static int mt6397_probe(struct platform_device *pdev)default :dev_err (&pdev->dev,"unsupported chip: %d\n" , id); - ret = -ENODEV ; -break ; +return -ENODEV ; }if (ret) {
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 6/34 mfd: ab8500-core: Return zero in get_register_interruptible()
In reply to Sasha LevinFrom: Dan Carpenter <[email protected]>
[ Upstream commit 10628e3ecf544fa2e4e24f8e112d95c37884dc98 ]
This function is supposed to return zero on success or negative error codes on error. Unfortunately, there is a bug so it sometimes returns non-zero, positive numbers on success.
I noticed this bug during review and I can't test it. It does appear that the return is sometimes propogated back to _regmap_read() where all non-zero returns are treated as failure so this may affect run time.
--- drivers/mfd/ab8500-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/mfd/ab8500-core.c +1 -1
@@ -261,7 +261,7 @@ static int get_register_interruptible(struct ab8500 *ab8500, u8 bank,mutex_unlock (&ab8500->lock);dev_vdbg (ab8500->dev,"rd: addr %#x => data %#x\n" , addr, ret); -return ret; +return (ret <0 ) ? ret :0 ; } static int ab8500_get_register(struct device *dev, u8 bank,
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 7/34 mfd: bd9571mwv: Add volatile register to make DVFS work
In reply to Sasha LevinFrom: Dien Pham <[email protected]>
[ Upstream commit b0aff01e7aa6ad2d6998ef1323843212d1db8b04 ]
Because BD9571MWV_DVFS_MONIVDAC is not defined in the volatile table, the physical register value is not updated by regmap and DVFS doesn't work as expected. Fix it!
--- drivers/mfd/bd9571mwv.c | 1 + 1 file changed, 1 insertion(+)
drivers/mfd/bd9571mwv.c +1 -0
@@ -57,6 +57,7 @@ static const struct regmap_access_table bd9571mwv_writable_table = { };static const struct regmap_range bd9571mwv_volatile_yes_ranges[] = { +regmap_reg_range (BD9571MWV_DVFS_MONIVDAC ,BD9571MWV_DVFS_MONIVDAC ),regmap_reg_range (BD9571MWV_GPIO_IN ,BD9571MWV_GPIO_IN ),regmap_reg_range (BD9571MWV_GPIO_INT ,BD9571MWV_GPIO_INT ),regmap_reg_range (BD9571MWV_INT_INTREQ ,BD9571MWV_INT_INTREQ ),
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 8/34 mfd: qcom_rpm: write fw_version to CTRL_REG
In reply to Sasha LevinFrom: Jonathan Marek <[email protected]>
[ Upstream commit 504e4175829c44328773b96ad9c538e4783a8d22 ]
This is required as part of the initialization sequence on certain SoCs.
If these registers are not initialized, the hardware can be unresponsive. This fixes the driver on apq8060 (HP TouchPad device).
--- drivers/mfd/qcom_rpm.c | 4 ++++ 1 file changed, 4 insertions(+)
drivers/mfd/qcom_rpm.c +4 -0
@@ -638,6 +638,10 @@ static int qcom_rpm_probe(struct platform_device *pdev)return -EFAULT ; } +writel (fw_version[0 ],RPM_CTRL_REG (rpm,0 )); +writel (fw_version[1 ],RPM_CTRL_REG (rpm,1 )); +writel (fw_version[2 ],RPM_CTRL_REG (rpm,2 )); +dev_info (&pdev->dev,"RPM firmware %u.%u.%u\n" , fw_version[0 ], fw_version[1 ], fw_version[2 ]);
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 9/34 mfd: wm5110: Add missing ASRC rate register
In reply to Sasha LevinFrom: Charles Keepax <[email protected]>
[ Upstream commit 04c801c18ded421845324255e660147a6f58dcd6 ]
--- drivers/mfd/wm5110-tables.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/mfd/wm5110-tables.c +2 -0
@@ -1618,6 +1618,7 @@ static const struct reg_default wm5110_reg_default[] = { { 0x00000ECD, 0x0000 }, /* R3789 - HPLPF4_2 */ { 0x00000EE0, 0x0000 }, /* R3808 - ASRC_ENABLE */ { 0x00000EE2, 0x0000 }, /* R3810 - ASRC_RATE1 */ + { 0x00000EE3, 0x4000 }, /* R3811 - ASRC_RATE2 */ { 0x00000EF0, 0x0000 }, /* R3824 - ISRC 1 CTRL 1 */ { 0x00000EF1, 0x0000 }, /* R3825 - ISRC 1 CTRL 2 */ { 0x00000EF2, 0x0000 }, /* R3826 - ISRC 1 CTRL 3 */
@@ -2869,6 +2870,7 @@ static bool wm5110_readable_register(struct device *dev, unsigned int reg)case ARIZONA_ASRC_ENABLE :case ARIZONA_ASRC_STATUS :case ARIZONA_ASRC_RATE1 : +case ARIZONA_ASRC_RATE2 :case ARIZONA_ISRC_1_CTRL_1 :case ARIZONA_ISRC_1_CTRL_2 :case ARIZONA_ISRC_1_CTRL_3 :
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 10/34 mfd: tps65218: Use devm_regmap_add_irq_chip and clean up error path in probe()
In reply to Sasha LevinFrom: Keerthy <[email protected]>
[ Upstream commit 75d4c5e03c2ae9902ab521024b10291f6fc9515b ]
Use devm_regmap_add_irq_chip and clean up error path in probe and also the remove function.
--- drivers/mfd/tps65218.c | 24 +++--------------------- 1 file changed, 3 insertions(+), 21 deletions(-)
drivers/mfd/tps65218.c +3 -21
@@ -243,9 +243,9 @@ static int tps65218_probe(struct i2c_client *client,mutex_init (&tps->tps_lock); - ret =regmap_add_irq_chip (tps->regmap, tps->irq, -IRQF_ONESHOT ,0 , &tps65218_irq_chip, - &tps->irq_data); + ret =devm_regmap_add_irq_chip (&client->dev, tps->regmap, tps->irq, +IRQF_ONESHOT ,0 , &tps65218_irq_chip, + &tps->irq_data);if (ret <0 )return ret;
@@ -261,26 +261,9 @@ static int tps65218_probe(struct i2c_client *client,ARRAY_SIZE (tps65218_cells),NULL ,0 , regmap_irq_get_domain(tps->irq_data)); -if (ret <0 ) - goto err_irq; - -return 0 ; - -err_irq: -regmap_del_irq_chip (tps->irq, tps->irq_data); -return ret; } -static int tps65218_remove (struct i2c_client *client) -{ -struct tps65218 *tps =i2c_get_clientdata (client); - -regmap_del_irq_chip (tps->irq, tps->irq_data); - -return 0 ; -} -static const struct i2c_device_id tps65218_id_table[] = { {"tps65218" ,TPS65218 }, { },
@@ -293,7 +276,6 @@ static struct i2c_driver tps65218_driver = { .of_match_table = of_tps65218_match_table, }, .probe = tps65218_probe, - .remove = tps65218_remove, .id_table = tps65218_id_table, };
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 11/34 mfd: mc13xxx: Fix a missing check of a register-read failure
In reply to Sasha LevinFrom: Kangjie Lu <[email protected]>
[ Upstream commit 9e28989d41c0eab57ec0bb156617a8757406ff8a ]
When mc13xxx_reg_read() fails, "old_adc0" is uninitialized and will contain random value. Further execution uses "old_adc0" even when mc13xxx_reg_read() fails.
The fix checks the return value of mc13xxx_reg_read(), and exits the execution when it fails.
--- drivers/mfd/mc13xxx-core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/mfd/mc13xxx-core.c +3 -1
@@ -274,7 +274,9 @@ int mc13xxx_adc_do_conversion(struct mc13xxx *mc13xxx, unsigned int mode, mc13xxx->adcflags |=MC13XXX_ADC_WORKING ; -mc13xxx_reg_read (mc13xxx,MC13XXX_ADC0 , &old_adc0); + ret =mc13xxx_reg_read (mc13xxx,MC13XXX_ADC0 , &old_adc0); +if (ret) + goto out; adc0 =MC13XXX_ADC0_ADINC1 |MC13XXX_ADC0_ADINC2 ; adc1 =MC13XXX_ADC1_ADEN |MC13XXX_ADC1_ADTRIGIGN |MC13XXX_ADC1_ASC ;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 12/34 xen/pvcalls: remove set but not used variable 'intf'
In reply to Sasha LevinFrom: YueHaibing <[email protected]>
[ Upstream commit 1f8ce09b36c41a026a37a24b20efa32000892a64 ]
Fixes gcc '-Wunused-but-set-variable' warning:
drivers/xen/pvcalls-back.c: In function 'pvcalls_sk_state_change':drivers/xen/pvcalls-back.c:286:28: warning:
variable 'intf' set but not used [-Wunused-but-set-variable]
It not used since e6587cdbd732 ("pvcalls-back: set -ENOTCONN in pvcalls_conn_back_read")
--- drivers/xen/pvcalls-back.c | 2 -- 1 file changed, 2 deletions(-)
drivers/xen/pvcalls-back.c +0 -2
@@ -291,12 +291,10 @@ static int pvcalls_back_socket(struct xenbus_device *dev,static void pvcalls_sk_state_change (struct sock *sock) {struct sock_mapping *map = sock->sk_user_data; -struct pvcalls_data_intf *intf;if (map ==NULL )return ; - intf = map->ring;atomic_inc (&map->read);notify_remote_via_irq (map->irq); }
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 13/34 qed: Fix qed_chain_set_prod() for PBL chains with non power of 2 page count
In reply to Sasha LevinFrom: Denis Bolotin <[email protected]>
[ Upstream commit 2d533a9287f2011632977e87ce2783f4c689c984 ]
In PBL chains with non power of 2 page count, the producer is not at the beginning of the chain when index is 0 after a wrap. Therefore, after the producer index wrap around, page index should be calculated more carefully.
--- include/linux/qed/qed_chain.h | 31 +++++++++++++++++++++++++++++++ 1 file changed, 31 insertions(+)
include/linux/qed/qed_chain.h +31 -0
@@ -663,6 +663,37 @@ static inline void *qed_chain_get_last_elem(struct qed_chain *p_chain)static inline void qed_chain_set_prod (struct qed_chain *p_chain,u32 prod_idx,void *p_prod_elem) { +if (p_chain->mode ==QED_CHAIN_MODE_PBL ) { +u32 cur_prod, page_mask, page_cnt, page_diff; + + cur_prod =is_chain_u16 (p_chain) ? p_chain->u.chain16.prod_idx : + p_chain->u.chain32.prod_idx; + +/* Assume that number of elements in a page is power of 2 */ + page_mask = ~p_chain->elem_per_page_mask; + +/* Use "cur_prod - 1" and "prod_idx - 1" since producer index +* reaches the first element of next page before the page index +* is incremented. See qed_chain_produce(). +* Index wrap around is not a problem because the difference +* between current and given producer indices is always +* positive and lower than the chain's capacity. +*/ + page_diff = (((cur_prod -1 ) & page_mask) - + ((prod_idx -1 ) & page_mask)) / + p_chain->elem_per_page; + + page_cnt =qed_chain_get_page_cnt (p_chain); +if (is_chain_u16 (p_chain)) + p_chain->pbl.c.u16.prod_page_idx = + (p_chain->pbl.c.u16.prod_page_idx - + page_diff + page_cnt) % page_cnt; +else + p_chain->pbl.c.u32.prod_page_idx = + (p_chain->pbl.c.u32.prod_page_idx - + page_diff + page_cnt) % page_cnt; + } +if (is_chain_u16 (p_chain)) p_chain->u.chain16.prod_idx = (u16 ) prod_idx;else
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 14/34 qed: Fix qed_ll2_post_rx_buffer_notify_fw() by adding a write memory barrier
In reply to Sasha LevinFrom: Denis Bolotin <[email protected]>
[ Upstream commit 46721c3d9e273aea880e9ff835b0e1271e1cd2fb ]
Make sure chain element is updated before ringing the doorbell.
--- drivers/net/ethernet/qlogic/qed/qed_ll2.c | 4 ++++ 1 file changed, 4 insertions(+)
drivers/net/ethernet/qlogic/qed/qed_ll2.c +4 -0
@@ -1487,6 +1487,10 @@ static void qed_ll2_post_rx_buffer_notify_fw(struct qed_hwfn *p_hwfn, cq_prod =qed_chain_get_prod_idx (&p_rx->rcq_chain); rx_prod.bd_prod =cpu_to_le16 (bd_prod); rx_prod.cqe_prod =cpu_to_le16 (cq_prod); + +/* Make sure chain element is updated before ringing the doorbell */ +dma_wmb (); +DIRECT_REG_WR (p_rx->set_prod_addr, *((u32 *)&rx_prod)); }
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 15/34 net: hns: Fix use after free identified by SLUB debug
In reply to Sasha LevinFrom: Yonglong Liu <[email protected]>
[ Upstream commit bb989501abcafa0de5f18b0ec0ec459b5b817908 ]
When enable SLUB debug, than remove hns_enet_drv module, SLUB debug will identify a use after free bug:
[134.189505] Unable to handle kernel paging request at virtual address 006b6b6b6b6b6b6b [134.197553] Mem abort info: [134.200381] ESR = 0x96000004 [134.203487] Exception class = DABT (current EL), IL = 32 bits [134.209497] SET = 0, FnV = 0 [134.212596] EA = 0, S1PTW = 0 [134.215777] Data abort info: [134.218701] ISV = 0, ISS = 0x00000004 [134.222596] CM = 0, WnR = 0 [134.225606] [006b6b6b6b6b6b6b] address between user and kernel address ranges [134.232851] Internal error: Oops: 96000004 [#1] SMP [134.237798] CPU: 21 PID: 27834 Comm: rmmod Kdump: loaded Tainted: G OE 4.19.5-1.2.34.aarch64 #1 [134.247856] Hardware name: Huawei TaiShan 2280 /BC11SPCD, BIOS 1.58 10/24/2018 [134.255181] pstate: 20000005 (nzCv daif -PAN -UAO) [134.260044] pc : hns_ae_put_handle+0x38/0x60 [134.264372] lr : hns_ae_put_handle+0x24/0x60 [134.268700] sp : ffff00001be93c50 [134.272054] x29: ffff00001be93c50 x28: ffff802faaec8040 [134.277442] x27: 0000000000000000 x26: 0000000000000000 [134.282830] x25: 0000000056000000 x24: 0000000000000015 [134.288284] x23: ffff0000096fe098 x22: ffff000001050070 [134.293671] x21: ffff801fb3c044a0 x20: ffff80afb75ec098 [134.303287] x19: ffff80afb75ec098 x18: 0000000000000000 [134.312945] x17: 0000000000000000 x16: 0000000000000000 [134.322517] x15: 0000000000000002 x14: 0000000000000000 [134.332030] x13: dead000000000100 x12: ffff7e02bea3c988 [134.341487] x11: ffff80affbee9e68 x10: 0000000000000000 [134.351033] x9 : 6fffff8000008101 x8 : 0000000000000000 [134.360569] x7 : dead000000000100 x6 : ffff000009579748 [134.370059] x5 : 0000000000210d00 x4 : 0000000000000000 [134.379550] x3 : 0000000000000001 x2 : 0000000000000000 [134.388813] x1 : 6b6b6b6b6b6b6b6b x0 : 0000000000000000 [134.397993] Process rmmod (pid: 27834, stack limit = 0x00000000d474b7fd) [134.408498] Call trace: [134.414611] hns_ae_put_handle+0x38/0x60 [134.422208] hnae_put_handle+0xd4/0x108 [134.429563] hns_nic_dev_remove+0x60/0xc0 [hns_enet_drv] [134.438342] platform_drv_remove+0x2c/0x70 [134.445958] device_release_driver_internal+0x174/0x208 [134.454810] driver_detach+0x70/0xd8 [134.461913] bus_remove_driver+0x64/0xe8 [134.469396] driver_unregister+0x34/0x60 [134.476822] platform_driver_unregister+0x20/0x30 [134.485130] hns_nic_dev_driver_exit+0x14/0x6e4 [hns_enet_drv] [134.494634] __arm64_sys_delete_module+0x238/0x290
struct hnae_handle is a member of struct hnae_vf_cb, so when vf_cb is freed, than use hnae_handle will cause use after free panic.
This patch frees vf_cb after hnae_handle used.
--- drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/net/ethernet/hisilicon/hns/hns_ae_adapt.c +2 -4
@@ -147,12 +147,10 @@ static void hns_ae_put_handle(struct hnae_handle *handle)struct hnae_vf_cb *vf_cb =hns_ae_get_vf_cb (handle);int i; - vf_cb->mac_cb =NULL ; - -kfree (vf_cb); -for (i =0 ; i < handle->q_num; i++)hns_ae_get_ring_pair (handle->qs[i])->used_by_vf =0 ; + +kfree (vf_cb); }static int hns_ae_wait_flow_down (struct hnae_handle *handle)
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 16/34 MIPS: ath79: Enable OF serial ports in the default config
In reply to Sasha LevinFrom: Alban Bedel <[email protected]>
[ Upstream commit 565dc8a4f55e491935bfb04866068d21784ea9a4 ]
CONFIG_SERIAL_OF_PLATFORM is needed to get a working console on the OF boards, enable it in the default config to get a working setup out of the box.
--- arch/mips/configs/ath79_defconfig | 1 + 1 file changed, 1 insertion(+)
arch/mips/configs/ath79_defconfig +1 -0
@@ -72,6 +72,7 @@ CONFIG_SERIAL_8250_CONSOLE=y # CONFIG_SERIAL_8250_PCI is not set CONFIG_SERIAL_8250_NR_UARTS=1 CONFIG_SERIAL_8250_RUNTIME_UARTS=1 +CONFIG_SERIAL_OF_PLATFORM=y CONFIG_SERIAL_AR933X=y CONFIG_SERIAL_AR933X_CONSOLE=y # CONFIG_HW_RANDOM is not set
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 17/34 netfilter: nf_tables: fix leaking object reference count
In reply to Sasha LevinFrom: Taehee Yoo <[email protected]>
[ Upstream commit b91d9036883793122cf6575ca4dfbfbdd201a83d ]
There is no code that decreases the reference count of stateful objects in error path of the nft_add_set_elem(). this causes a leak of reference count of stateful objects.
Test commands:
$nft add table ip filter
$nft add counter ip filter c1
$nft add map ip filter m1 { type ipv4_addr : counter \;} $nft add element ip filter m1 { 1 : c1 }
$nft add element ip filter m1 { 1 : c1 }
$nft delete element ip filter m1 { 1 }
$nft delete counter ip filter c1
Result:
Error: Could not process rule: Device or resource busy delete counter ip filter c1 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^
At the second 'nft add element ip filter m1 { 1 : c1 }', the reference count of the 'c1' is increased then it tries to insert into the 'm1'. but the 'm1' already has same element so it returns -EEXIST.
But it doesn't decrease the reference count of the 'c1' in the error path. Due to a leak of the reference count of the 'c1', the 'c1' can't be removed by 'nft delete counter ip filter c1'.
--- net/netfilter/nf_tables_api.c | 2 ++ 1 file changed, 2 insertions(+)
net/netfilter/nf_tables_api.c +2 -0
@@ -4046,6 +4046,8 @@ static int nft_add_set_elem(struct nft_ctx *ctx, struct nft_set *set, err5:kfree (trans); err4: +if (obj) + obj->use--;kfree (elem.priv); err3:if (nla[NFTA_SET_ELEM_DATA ] !=NULL )
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 18/34 scsi: qla4xxx: check return code of qla4xxx_copy_from_fwddb_param
In reply to Sasha LevinFrom: YueHaibing <[email protected]>
[ Upstream commit 72b4a0465f995175a2e22cf4a636bf781f1f28a7 ]
The return code should be check while qla4xxx_copy_from_fwddb_param fails.
--- drivers/scsi/qla4xxx/ql4_os.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/scsi/qla4xxx/ql4_os.c +2 -0
@@ -7241,6 +7241,8 @@ static int qla4xxx_sysfs_ddb_tgt_create(struct scsi_qla_host *ha, rc =qla4xxx_copy_from_fwddb_param (fnode_sess, fnode_conn, fw_ddb_entry); +if (rc) + goto free_sess;ql4_printk (KERN_INFO , ha,"%s: sysfs entry %s created\n" , __func__, fnode_sess->dev.kobj.name);
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 19/34 scsi: isci: initialize shost fully before calling scsi_add_host()
In reply to Sasha LevinFrom: Logan Gunthorpe <[email protected]>
[ Upstream commit cc29a1b0a3f2597ce887d339222fa85b9307706d ]
scsi_mq_setup_tags(), which is called by scsi_add_host(), calculates the command size to allocate based on the prot_capabilities. In the isci driver, scsi_host_set_prot() is called after scsi_add_host() so the command size gets calculated to be smaller than it needs to be. Eventually, scsi_mq_init_request() locates the 'prot_sdb' after the command assuming it was sized correctly and a buffer overrun may occur.
However, seeing blk_mq_alloc_rqs() rounds up to the nearest cache line size, the mistake can go unnoticed.
The bug was noticed after the struct request size was reduced by commit 9d037ad707ed ("block: remove req->timeout_list")
Which likely reduced the allocated space for the request by an entire cache line, enough that the overflow could be hit and it caused a panic, on boot, at:
RIP: 0010:t10_pi_complete+0x77/0x1c0
Call Trace:
<IRQ>
sd_done+0xf5/0x340
scsi_finish_command+0xc3/0x120
blk_done_softirq+0x83/0xb0
__do_softirq+0xa1/0x2e6
irq_exit+0xbc/0xd0
call_function_single_interrupt+0xf/0x20
</IRQ>sd_done() would call scsi_prot_sg_count() which reads the number of entities in 'prot_sdb', but seeing 'prot_sdb' is located after the end of the allocated space it reads a garbage number and erroneously calls t10_pi_complete().
To prevent this, the calls to scsi_host_set_prot() are moved into isci_host_alloc() before the call to scsi_add_host(). Out of caution, also move the similar call to scsi_host_set_guard().
--- drivers/scsi/isci/init.c | 14 +++++++------- 1 file changed, 7 insertions(+), 7 deletions(-)
drivers/scsi/isci/init.c +7 -7
@@ -591,6 +591,13 @@ static struct isci_host *isci_host_alloc(struct pci_dev *pdev, int id) shost->max_lun = ~0 ; shost->max_cmd_len =MAX_COMMAND_SIZE ; +/* turn on DIF support */ +scsi_host_set_prot (shost, +SHOST_DIF_TYPE1_PROTECTION | +SHOST_DIF_TYPE2_PROTECTION | +SHOST_DIF_TYPE3_PROTECTION ); +scsi_host_set_guard (shost,SHOST_DIX_GUARD_CRC ); + err =scsi_add_host (shost, &pdev->dev);if (err) goto err_shost;
@@ -678,13 +685,6 @@ static int isci_pci_probe(struct pci_dev *pdev, const struct pci_device_id *id) goto err_host_alloc; } pci_info->hosts[i] = h; - -/* turn on DIF support */ -scsi_host_set_prot (to_shost (h), -SHOST_DIF_TYPE1_PROTECTION | -SHOST_DIF_TYPE2_PROTECTION | -SHOST_DIF_TYPE3_PROTECTION ); -scsi_host_set_guard (to_shost (h),SHOST_DIX_GUARD_CRC ); } err =isci_setup_interrupts (pdev);
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 20/34 MIPS: jazz: fix 64bit build
In reply to Sasha LevinFrom: Thomas Bogendoerfer <[email protected]>
[ Upstream commit 41af167fbc0032f9d7562854f58114eaa9270336 ]
64bit JAZZ builds failed with
linux-next/arch/mips/jazz/jazzdma.c: In function vdma_init:
/linux-next/arch/mips/jazz/jazzdma.c:77:30: error: implicit declaration
of function `KSEG1ADDR`; did you mean `CKSEG1ADDR`?
[-Werror=implicit-function-declaration]
pgtbl = (VDMA_PGTBL_ENTRY *)KSEG1ADDR(pgtbl);
^~~~~~~~~
CKSEG1ADDR
/linux-next/arch/mips/jazz/jazzdma.c:77:10: error: cast to pointer from
integer of different size [-Werror=int-to-pointer-cast]
pgtbl = (VDMA_PGTBL_ENTRY *)KSEG1ADDR(pgtbl);
^
In file included from /linux-next/arch/mips/include/asm/barrier.h:11:0,
from /linux-next/include/linux/compiler.h:248,
from /linux-next/include/linux/kernel.h:10,
from /linux-next/arch/mips/jazz/jazzdma.c:11:
/linux-next/arch/mips/include/asm/addrspace.h:41:29: error: cast from
pointer to integer of different size [-Werror=pointer-to-int-cast]
#define _ACAST32_ (_ATYPE_)(_ATYPE32_) /* widen if necessary */
^
/linux-next/arch/mips/include/asm/addrspace.h:53:25: note: in
expansion of macro `_ACAST32_`
#define CPHYSADDR(a) ((_ACAST32_(a)) & 0x1fffffff)
^~~~~~~~~
/linux-next/arch/mips/jazz/jazzdma.c:84:44: note: in expansion of
macro `CPHYSADDR`
r4030_write_reg32(JAZZ_R4030_TRSTBL_BASE, CPHYSADDR(pgtbl));Using correct casts and CKSEG1ADDR when dealing with the pgtbl setup fixes this.
--- arch/mips/jazz/jazzdma.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-)
arch/mips/jazz/jazzdma.c +3 -2
@@ -72,14 +72,15 @@ static int __init vdma_init(void) get_order(VDMA_PGTBL_SIZE));BUG_ON (!pgtbl);dma_cache_wback_inv ((unsigned long )pgtbl,VDMA_PGTBL_SIZE ); - pgtbl = (VDMA_PGTBL_ENTRY *)KSEG1ADDR (pgtbl); + pgtbl = (VDMA_PGTBL_ENTRY *)CKSEG1ADDR ((unsigned long )pgtbl);/* * Clear the R4030 translation table */ vdma_pgtbl_init (); -r4030_write_reg32 (JAZZ_R4030_TRSTBL_BASE ,CPHYSADDR (pgtbl)); +r4030_write_reg32 (JAZZ_R4030_TRSTBL_BASE , +CPHYSADDR ((unsigned long )pgtbl));r4030_write_reg32 (JAZZ_R4030_TRSTBL_LIM ,VDMA_PGTBL_SIZE );r4030_write_reg32 (JAZZ_R4030_TRSTBL_INV ,0 );
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 21/34 bpf: correctly set initial window on active Fast Open sender
In reply to Sasha LevinFrom: Yuchung Cheng <[email protected]>
[ Upstream commit 31aa6503a15ba00182ea6dbbf51afb63bf9e851d ]
The existing BPF TCP initial congestion window (TCP_BPF_IW) does not to work on (active) Fast Open sender. This is because it changes the (initial) window only if data_segs_out is zero -- but data_segs_out is also incremented on SYN-data. This patch fixes the issue by proerly accounting for SYN-data additionally.
--- net/core/filter.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
net/core/filter.c +1 -1
@@ -3128,7 +3128,7 @@ BPF_CALL_5(bpf_setsockopt, struct bpf_sock_ops_kern *, bpf_sock,/* Only some options are supported */ switch (optname) {case TCP_BPF_IW : -if (val <=0 || tp->data_segs_out >0 ) +if (val <=0 || tp->data_segs_out > tp->syn_data) ret = -EINVAL ;else tp->snd_cwnd = val;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 22/34 net: stmmac: Fix PCI module removal leak
In reply to Sasha LevinFrom: Jose Abreu <[email protected]>
[ Upstream commit 6dea7e1881fd86b80da64e476ac398008daed857 ]
Since commit b7d0f08e9129, the enable / disable of PCI device is not managed which will result in IO regions not being automatically unmapped. As regions continue mapped it is currently not possible to remove and then probe again the PCI module of stmmac.
Fix this by manually unmapping regions on remove callback.
Changes from v1:
- Fix build error
--- drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
drivers/net/ethernet/stmicro/stmmac/stmmac_pci.c +10 -0
@@ -299,7 +299,17 @@ static int stmmac_pci_probe(struct pci_dev *pdev,*/ static void stmmac_pci_remove (struct pci_dev *pdev) { +int i; +stmmac_dvr_remove (&pdev->dev); + +for (i =0 ; i <=PCI_STD_RESOURCE_END ; i++) { +if (pci_resource_len (pdev, i) ==0 ) +continue ; +pcim_iounmap_regions (pdev,BIT (i)); +break ; + } +pci_disable_device (pdev); }
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 23/34 isdn: i4l: isdn_tty: Fix some concurrency double-free bugs
In reply to Sasha LevinFrom: Jia-Ju Bai <[email protected]>
[ Upstream commit 2ff33d6637393fe9348357285931811b76e1402f ]
The functions isdn_tty_tiocmset() and isdn_tty_set_termios() may be concurrently executed.
isdn_tty_tiocmset
isdn_tty_modem_hup
line 719: kfree(info->dtmf_state);
line 721: kfree(info->silence_state);
line 723: kfree(info->adpcms);
line 725: kfree(info->adpcmr);isdn_tty_set_termios
isdn_tty_modem_hup
line 719: kfree(info->dtmf_state);
line 721: kfree(info->silence_state);
line 723: kfree(info->adpcms);
line 725: kfree(info->adpcmr);Thus, some concurrency double-free bugs may occur.
These possible bugs are found by a static tool written by myself and my manual code review.
To fix these possible bugs, the mutex lock "modem_info_mutex" used in isdn_tty_tiocmset() is added in isdn_tty_set_termios().
--- drivers/isdn/i4l/isdn_tty.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/isdn/i4l/isdn_tty.c +5 -1
@@ -1456,15 +1456,19 @@ isdn_tty_set_termios(struct tty_struct *tty, struct ktermios *old_termios) {modem_info *info = (modem_info *) tty->driver_data; +mutex_lock (&modem_info_mutex);if (!old_termios)isdn_tty_change_speed (info);else {if (tty->termios.c_cflag == old_termios->c_cflag && tty->termios.c_ispeed == old_termios->c_ispeed && - tty->termios.c_ospeed == old_termios->c_ospeed) + tty->termios.c_ospeed == old_termios->c_ospeed) { +mutex_unlock (&modem_info_mutex);return ; + }isdn_tty_change_speed (info); } +mutex_unlock (&modem_info_mutex); }/*
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 24/34 scsi: ufs: Fix system suspend status
In reply to Sasha LevinFrom: Stanley Chu <[email protected]>
[ Upstream commit ce9e7bce43526626f7cffe2e657953997870197e ]
hba->is_sys_suspended is set after successful system suspend but not clear after successful system resume.
According to current behavior, hba->is_sys_suspended will not be set if host is runtime-suspended but not system-suspended. Thus we shall aligh the same policy: clear this flag even if host remains runtime-suspended after ufshcd_system_resume is successfully returned.
Simply fix this flag to correct host status logs.
--- drivers/scsi/ufs/ufshcd.c | 2 ++ 1 file changed, 2 insertions(+)
drivers/scsi/ufs/ufshcd.c +2 -0
@@ -7520,6 +7520,8 @@ int ufshcd_system_resume(struct ufs_hba *hba)trace_ufshcd_system_resume (dev_name (hba->dev), ret,ktime_to_us (ktime_sub (ktime_get (), start)), hba->curr_dev_pwr_mode, hba->uic_link_state); +if (!ret) + hba->is_sys_suspended = false;return ret; }EXPORT_SYMBOL (ufshcd_system_resume);
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 25/34 scsi: qedi: Add ep_state for login completion on un-reachable targets
In reply to Sasha LevinFrom: Manish Rangankar <[email protected]>
[ Upstream commit 34a2ce887668db9dda4b56e6f155c49ac13f3e54 ]
When the driver finds invalid destination MAC for the first un-reachable target, and before completes the PATH_REQ operation, set new ep_state to OFFLDCONN_NONE so that as part of driver ep_poll mechanism, the upper open-iscsi layer is notified to complete the login process on the first un-reachable target and thus proceed login to other reachable targets.
--- drivers/scsi/qedi/qedi_iscsi.c | 3 +++ drivers/scsi/qedi/qedi_iscsi.h | 1 + 2 files changed, 4 insertions(+)
drivers/scsi/qedi/qedi_iscsi.c +3 -0
@@ -961,6 +961,7 @@ static int qedi_ep_poll(struct iscsi_endpoint *ep, int timeout_ms) qedi_ep = ep->dd_data;if (qedi_ep->state ==EP_STATE_IDLE || + qedi_ep->state ==EP_STATE_OFLDCONN_NONE || qedi_ep->state ==EP_STATE_OFLDCONN_FAILED )return -1 ;
@@ -1043,6 +1044,7 @@ static void qedi_ep_disconnect(struct iscsi_endpoint *ep)switch (qedi_ep->state) {case EP_STATE_OFLDCONN_START : +case EP_STATE_OFLDCONN_NONE : goto ep_release_conn;case EP_STATE_OFLDCONN_FAILED :break ;
@@ -1233,6 +1235,7 @@ static int qedi_set_path(struct Scsi_Host *shost, struct iscsi_path *path_data)if (!is_valid_ether_addr (&path_data->mac_addr[0 ])) {QEDI_NOTICE (&qedi->dbg_ctx,"dst mac NOT VALID\n" ); + qedi_ep->state =EP_STATE_OFLDCONN_NONE ; ret = -EIO ; goto set_path_exit; }
drivers/scsi/qedi/qedi_iscsi.h +1 -0
@@ -59,6 +59,7 @@ enum {EP_STATE_OFLDCONN_FAILED =0x2000 ,EP_STATE_CONNECT_FAILED =0x4000 ,EP_STATE_DISCONN_TIMEDOUT =0x8000 , + EP_STATE_OFLDCONN_NONE = 0x10000, };struct qedi_conn ;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 26/34 always clear the X2APIC_ENABLE bit for PV guest
In reply to Sasha LevinFrom: Talons Lee <[email protected]>
[ Upstream commit 5268c8f39e0efef81af2aaed160272d9eb507beb ]
Commit e657fcc clears cpu capability bit instead of using fake cpuid value, the EXTD should always be off for PV guest without depending on cpuid value. So remove the cpuid check in xen_read_msr_safe() to always clear the X2APIC_ENABLE bit.
--- arch/x86/xen/enlighten_pv.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-)
arch/x86/xen/enlighten_pv.c +1 -4
@@ -900,10 +900,7 @@ static u64 xen_read_msr_safe(unsigned int msr, int *err) val =native_read_msr_safe (msr, err);switch (msr) {case MSR_IA32_APICBASE : -#ifdef CONFIG_X86_X2APIC - if (!(cpuid_ecx(1) & (1 << (X86_FEATURE_X2APIC & 31)))) -#endif - val &= ~X2APIC_ENABLE ; + val &= ~X2APIC_ENABLE ;break ; }return val;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 27/34 drm/meson: add missing of_node_put
In reply to Sasha LevinFrom: Julia Lawall <[email protected]>
[ Upstream commit f672b93e4a0a4947d2e1103ed8780e01e13eadb6 ]
Add an of_node_put when the result of of_graph_get_remote_port_parent is not available.
An of_node_put is also needed when meson_probe_remote completes. This was present at the recursive call, but not in the call from meson_drv_probe.
The semantic match that finds this problem is as follows
(http://coccinelle.lip6.fr):
// <smpl>
@r exists@
local idexpression e;
expression x;
@@
e = of_graph_get_remote_port_parent(...);
... when != x = e
when != true e == NULL
when != of_node_put(e)
when != of_fwnode_handle(e)(
return e;
|
*return ...;
)
// </smpl>
--- drivers/gpu/drm/meson/meson_drv.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/gpu/drm/meson/meson_drv.c +7 -2
@@ -345,8 +345,10 @@ static int meson_probe_remote(struct platform_device *pdev, remote_node =of_graph_get_remote_port_parent (ep);if (!remote_node || remote_node == parent ||/* Ignore parent endpoint */ - !of_device_is_available (remote_node)) + !of_device_is_available (remote_node)) { +of_node_put (remote_node);continue ; + } count +=meson_probe_remote (pdev, match, remote, remote_node);
@@ -365,10 +367,13 @@ static int meson_drv_probe(struct platform_device *pdev)for_each_endpoint_of_node (np, ep) { remote =of_graph_get_remote_port_parent (ep); -if (!remote || !of_device_is_available (remote)) +if (!remote || !of_device_is_available (remote)) { +of_node_put (remote);continue ; + } count +=meson_probe_remote (pdev, &match, np, remote); +of_node_put (remote); }if (count && !match)
Signature · 6 lines
2.19.1 _______________________________________________ dri-devel mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/dri-devel
Sasha Levin
maintainerPATCH 28/34 atm: he: fix sign-extension overflow on large shift
In reply to Sasha LevinFrom: Colin Ian King <[email protected]>
[ Upstream commit cb12d72b27a6f41325ae23a11033cf5fedfa1b97 ]
Shifting the 1 by exp by an int can lead to sign-extension overlow when exp is 31 since 1 is an signed int and sign-extending this result to an unsigned long long will set the upper 32 bits. Fix this by shifting an unsigned long.
Detected by cppcheck:
(warning) Shifting signed 32-bit value by 31 bits is undefined behaviour
--- drivers/atm/he.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/atm/he.c +1 -1
@@ -717,7 +717,7 @@ static int he_init_cs_block_rcm(struct he_dev *he_dev)instead of '/ 512', use '>> 9' to prevent a call to divdu3 on x86 platforms */ - rate_cps = (unsigned long long ) (1 << exp) * (man +512 ) >>9 ; + rate_cps = (unsigned long long ) (1UL << exp) * (man +512 ) >>9 ;if (rate_cps <10 ) rate_cps =10 ;/* 2.2.1 minimum payload rate is 10 cps */
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 29/34 hwmon: (tmp421) Correct the misspelling of the tmp442 compatible attribute in OF device ID table
In reply to Sasha LevinFrom: Cheng-Min Ao <[email protected]>
[ Upstream commit f422449b58548a41e98fc97b259a283718e527db ]
Correct a typo in OF device ID table
The last one should be 'ti,tmp442'
--- drivers/hwmon/tmp421.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/hwmon/tmp421.c +1 -1
@@ -88,7 +88,7 @@ static const struct of_device_id tmp421_of_match[] = { .data = (void *)2 }, { - .compatible = "ti,tmp422", + .compatible = "ti,tmp442", .data = (void *)3 }, { },
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 30/34 leds: lp5523: fix a missing check of return value of lp55xx_read
In reply to Sasha LevinFrom: Kangjie Lu <[email protected]>
[ Upstream commit 248b57015f35c94d4eae2fdd8c6febf5cd703900 ]
When lp55xx_read() fails, "status" is an uninitialized variable and thus may contain random value; using it leads to undefined behaviors.
The fix inserts a check for the return value of lp55xx_read: if it fails, returns with its error code.
--- drivers/leds/leds-lp5523.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/leds/leds-lp5523.c +3 -1
@@ -318,7 +318,9 @@ static int lp5523_init_program_engine(struct lp55xx_chip *chip)/* Let the programs run for couple of ms and check the engine status */ usleep_range (3000 ,6000 ); -lp55xx_read (chip,LP5523_REG_STATUS , &status); + ret =lp55xx_read (chip,LP5523_REG_STATUS , &status); +if (ret) +return ret; status &=LP5523_ENG_STATUS_MASK ;if (status !=LP5523_ENG_STATUS_MASK ) {
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 31/34 bpf: bpf_setsockopt: reset sock dst on SO_MARK changes
In reply to Sasha LevinFrom: Peter Oskolkov <[email protected]>
[ Upstream commit f4924f24da8c7ef64195096817f3cde324091d97 ]
In sock_setsockopt() (net/core/sock.h), when SO_MARK option is used to change sk_mark, sk_dst_reset(sk) is called. The same should be done in bpf_setsockopt().
--- net/core/filter.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-)
net/core/filter.c +4 -1
@@ -3102,7 +3102,10 @@ BPF_CALL_5(bpf_setsockopt, struct bpf_sock_ops_kern *, bpf_sock, sk->sk_rcvlowat = val ? :1 ;break ;case SO_MARK : - sk->sk_mark = val; +if (sk->sk_mark != val) { + sk->sk_mark = val; +sk_dst_reset (sk); + }break ;default : ret = -EINVAL ;
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 32/34 mlxsw: spectrum_switchdev: Do not treat static FDB entries as sticky
In reply to Sasha LevinFrom: Ido Schimmel <[email protected]>
[ Upstream commit 64254a2054611205798e6bde634639bc704573ac ]
The driver currently treats static FDB entries as both static and sticky. This is incorrect and prevents such entries from being roamed to a different port via learning.
Fix this by configuring static entries with ageing disabled and roaming enabled.
In net-next we can add proper support for the newly introduced 'sticky' flag.
--- .../net/ethernet/mellanox/mlxsw/spectrum_switchdev.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-)
drivers/net/ethernet/mellanox/mlxsw/spectrum_switchdev.c +6 -6
@@ -1111,7 +1111,7 @@ mlxsw_sp_bridge_port_fdb_flush(struct mlxsw_sp *mlxsw_sp,static enum mlxsw_reg_sfd_rec_policy mlxsw_sp_sfd_rec_policy (bool dynamic) {return dynamic ?MLXSW_REG_SFD_REC_POLICY_DYNAMIC_ENTRY_INGRESS : -MLXSW_REG_SFD_REC_POLICY_STATIC_ENTRY ; +MLXSW_REG_SFD_REC_POLICY_DYNAMIC_ENTRY_MLAG ; }static enum mlxsw_reg_sfd_op mlxsw_sp_sfd_op (bool adding)
@@ -1123,7 +1123,7 @@ static enum mlxsw_reg_sfd_op mlxsw_sp_sfd_op(bool adding)static int __mlxsw_sp_port_fdb_uc_op (struct mlxsw_sp *mlxsw_sp,u8 local_port,const char *mac,u16 fid,bool adding,enum mlxsw_reg_sfd_rec_action action, -bool dynamic) +enum mlxsw_reg_sfd_rec_policy policy) {char *sfd_pl;u8 num_rec;
@@ -1134,8 +1134,7 @@ static int __mlxsw_sp_port_fdb_uc_op(struct mlxsw_sp *mlxsw_sp, u8 local_port,return -ENOMEM ;mlxsw_reg_sfd_pack (sfd_pl,mlxsw_sp_sfd_op (adding),0 ); -mlxsw_reg_sfd_uc_pack (sfd_pl,0 ,mlxsw_sp_sfd_rec_policy (dynamic), - mac, fid, action, local_port); +mlxsw_reg_sfd_uc_pack (sfd_pl,0 , policy, mac, fid, action, local_port); num_rec =mlxsw_reg_sfd_num_rec_get (sfd_pl); err =mlxsw_reg_write (mlxsw_sp->core,MLXSW_REG (sfd), sfd_pl);if (err)
@@ -1154,7 +1153,8 @@ static int mlxsw_sp_port_fdb_uc_op(struct mlxsw_sp *mlxsw_sp, u8 local_port, bool dynamic) {return __mlxsw_sp_port_fdb_uc_op (mlxsw_sp, local_port, mac, fid, adding, -MLXSW_REG_SFD_REC_ACTION_NOP , dynamic); +MLXSW_REG_SFD_REC_ACTION_NOP , +mlxsw_sp_sfd_rec_policy (dynamic)); } int mlxsw_sp_rif_fdb_op(struct mlxsw_sp *mlxsw_sp, const char *mac, u16 fid,
@@ -1162,7 +1162,7 @@ int mlxsw_sp_rif_fdb_op(struct mlxsw_sp *mlxsw_sp, const char *mac, u16 fid, {return __mlxsw_sp_port_fdb_uc_op (mlxsw_sp,0 , mac, fid, adding,MLXSW_REG_SFD_REC_ACTION_FORWARD_IP_ROUTER , - false); +MLXSW_REG_SFD_REC_POLICY_STATIC_ENTRY ); } static int mlxsw_sp_port_fdb_uc_lag_op(struct mlxsw_sp *mlxsw_sp, u16 lag_id,
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 33/34 net/mlx5e: Fix wrong (zero) TX drop counter indication for representor
In reply to Sasha LevinFrom: Tariq Toukan <[email protected]>
[ Upstream commit 7fdc1adc52d3975740547a78c2df329bb207f15d ]
For representors, the TX dropped counter is not folded from the per-ring counters. Fix it.
--- drivers/net/ethernet/mellanox/mlx5/core/en_rep.c | 1 + 1 file changed, 1 insertion(+)
drivers/net/ethernet/mellanox/mlx5/core/en_rep.c +1 -0
@@ -125,6 +125,7 @@ static void mlx5e_rep_update_sw_counters(struct mlx5e_priv *priv) s->tx_packets += sq_stats->packets; s->tx_bytes += sq_stats->bytes; + s->tx_queue_dropped += sq_stats->dropped; } } }
Signature · 1 line
2.19.1
Sasha Levin
maintainerPATCH 34/34 isdn: avm: Fix string plus integer warning from Clang
In reply to Sasha LevinFrom: Nathan Chancellor <[email protected]>
[ Upstream commit 7afa81c55fca0cad589722cb4bce698b4803b0e1 ]
A recent commit in Clang expanded the -Wstring-plus-int warning, showing some odd behavior in this file.
drivers/isdn/hardware/avm/b1.c:426:30: warning: adding 'int' to a string does not append to the string [-Wstring-plus-int]
cinfo->version[j] = "\0\0" + 1;
~~~~~~~^~~
drivers/isdn/hardware/avm/b1.c:426:30: note: use array indexing to silence this warning
cinfo->version[j] = "\0\0" + 1;
^
& [ ]1 warning generated.
This is equivalent to just "\0". Nick pointed out that it is smarter to use "" instead of "\0" because "" is used elsewhere in the kernel and can be deduplicated at the linking stage.
--- drivers/isdn/hardware/avm/b1.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/isdn/hardware/avm/b1.c +1 -1
@@ -423,7 +423,7 @@ void b1_parse_version(avmctrl_info *cinfo)int i, j;for (j =0 ; j <AVM_MAXVERSION ; j++) - cinfo->version[j] ="\0\0" +1 ; + cinfo->version[j] ="" ;for (i =0 , j =0 ; j <AVM_MAXVERSION && i < cinfo->versionlen; j++, i += cinfo->versionbuf[i] +1 )
Signature · 1 line
2.19.1