December 2024 - Linux-stable-mirror

[PATCH RFT 1/6] serial: sh-sci: Check if TX data was written to device in .tx_empty()

by Claudiu

From: Claudiu Beznea <claudiu.beznea.uj(a)bp.renesas.com> On the Renesas RZ/G3S, when doing suspend to RAM, the uart_suspend_port() is called. The uart_suspend_port() calls 3 times the struct uart_port::ops::tx_empty() before shutting down the port. According to the documentation, the struct uart_port::ops::tx_empty() API tests whether the transmitter FIFO and shifter for the port is empty. The Renesas RZ/G3S SCIFA IP reports the number of data units stored in the transmit FIFO through the FDR (FIFO Data Count Register). The data units in the FIFOs are written in the shift register and transmitted from there. The TEND bit in the Serial Status Register reports if the data was transmitted from the shift register. In the previous code, in the tx_empty() API implemented by the sh-sci driver, it is considered that the TX is empty if the hardware reports the TEND bit set and the number of data units in the FIFO is zero. According to the HW manual, the TEND bit has the following meaning: 0: Transmission is in the waiting state or in progress. 1: Transmission is completed. It has been noticed that when opening the serial device w/o using it and then switch to a power saving mode, the tx_empty() call in the uart_port_suspend() function fails, leading to the "Unable to drain transmitter" message being printed on the console. This is because the TEND=0 if nothing has been transmitted and the FIFOs are empty. As the TEND=0 has double meaning (waiting state, in progress) we can't determined the scenario described above. Add a software workaround for this. This sets a variable if any data has been sent on the serial console (when using PIO) or if the DMA callback has been called (meaning something has been transmitted). In the tx_empty() API the status of the DMA transaction is also checked and if it is completed or in progress the code falls back in checking the hardware registers instead of relying on the software variable. Fixes: 73a19e4c0301 ("serial: sh-sci: Add DMA support.") Cc: stable(a)vger.kernel.org Signed-off-by: Claudiu Beznea <claudiu.beznea.uj(a)bp.renesas.com> --- drivers/tty/serial/sh-sci.c | 29 +++++++++++++++++++++++++++++ 1 file changed, 29 insertions(+) diff --git a/drivers/tty/serial/sh-sci.c b/drivers/tty/serial/sh-sci.c index df523c744423..924b803af440 100644 --- a/drivers/tty/serial/sh-sci.c +++ b/drivers/tty/serial/sh-sci.c @@ -157,6 +157,7 @@ struct sci_port { bool has_rtscts; bool autorts; + bool tx_occurred; }; #define SCI_NPORTS CONFIG_SERIAL_SH_SCI_NR_UARTS @@ -850,6 +851,7 @@ static void sci_transmit_chars(struct uart_port *port) { struct tty_port *tport = &port->state->port; unsigned int stopped = uart_tx_stopped(port); + struct sci_port *s = to_sci_port(port); unsigned short status; unsigned short ctrl; int count; @@ -885,6 +887,7 @@ static void sci_transmit_chars(struct uart_port *port) } sci_serial_out(port, SCxTDR, c); + s->tx_occurred = true; port->icount.tx++; } while (--count > 0); @@ -1241,6 +1244,8 @@ static void sci_dma_tx_complete(void *arg) if (kfifo_len(&tport->xmit_fifo) < WAKEUP_CHARS) uart_write_wakeup(port); + s->tx_occurred = true; + if (!kfifo_is_empty(&tport->xmit_fifo)) { s->cookie_tx = 0; schedule_work(&s->work_tx); @@ -1731,6 +1736,19 @@ static void sci_flush_buffer(struct uart_port *port) s->cookie_tx = -EINVAL; } } + +static void sci_dma_check_tx_occurred(struct sci_port *s) +{ + struct dma_tx_state state; + enum dma_status status; + + if (!s->chan_tx) + return; + + status = dmaengine_tx_status(s->chan_tx, s->cookie_tx, &state); + if (status == DMA_COMPLETE || status == DMA_IN_PROGRESS) + s->tx_occurred = true; +} #else /* !CONFIG_SERIAL_SH_SCI_DMA */ static inline void sci_request_dma(struct uart_port *port) { @@ -1740,6 +1758,10 @@ static inline void sci_free_dma(struct uart_port *port) { } +static void sci_dma_check_tx_occurred(struct sci_port *s) +{ +} + #define sci_flush_buffer NULL #endif /* !CONFIG_SERIAL_SH_SCI_DMA */ @@ -2076,6 +2098,12 @@ static unsigned int sci_tx_empty(struct uart_port *port) { unsigned short status = sci_serial_in(port, SCxSR); unsigned short in_tx_fifo = sci_txfill(port); + struct sci_port *s = to_sci_port(port); + + sci_dma_check_tx_occurred(s); + + if (!s->tx_occurred) + return TIOCSER_TEMT; return (status & SCxSR_TEND(port)) && !in_tx_fifo ? TIOCSER_TEMT : 0; } @@ -2247,6 +2275,7 @@ static int sci_startup(struct uart_port *port) dev_dbg(port->dev, "%s(%d)\n", __func__, port->line); + s->tx_occurred = false; sci_request_dma(port); ret = sci_request_irq(s); -- 2.39.2

6 months, 1 week

3
2
0 0

+ vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep.patch added to mm-hotfixes-unstable branch

by Andrew Morton

The patch titled Subject: vmstat: disable vmstat_work on vmstat_cpu_down_prep() has been added to the -mm mm-hotfixes-unstable branch. Its filename is vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep.patch This patch will shortly appear at https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patche… This patch will later appear in the mm-hotfixes-unstable branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's *** Remember to use Documentation/process/submit-checklist.rst when testing your code *** The -mm tree is included into linux-next via the mm-everything branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm and is updated there every 2-3 working days ------------------------------------------------------ From: Koichiro Den <koichiro.den(a)canonical.com> Subject: vmstat: disable vmstat_work on vmstat_cpu_down_prep() Date: Sat, 21 Dec 2024 12:33:20 +0900 Even after mm/vmstat:online teardown, shepherd may still queue work for the dying cpu until the cpu is removed from online mask. While it's quite rare, this means that after unbind_workers() unbinds a per-cpu kworker, it potentially runs vmstat_update for the dying CPU on an irrelevant cpu before entering atomic AP states. When CONFIG_DEBUG_PREEMPT=y, it results in the following error with the backtrace. BUG: using smp_processor_id() in preemptible [00000000] code: \ kworker/7:3/1702 caller is refresh_cpu_vm_stats+0x235/0x5f0 CPU: 0 UID: 0 PID: 1702 Comm: kworker/7:3 Tainted: G Tainted: [N]=TEST Workqueue: mm_percpu_wq vmstat_update Call Trace: <TASK> dump_stack_lvl+0x8d/0xb0 check_preemption_disabled+0xce/0xe0 refresh_cpu_vm_stats+0x235/0x5f0 vmstat_update+0x17/0xa0 process_one_work+0x869/0x1aa0 worker_thread+0x5e5/0x1100 kthread+0x29e/0x380 ret_from_fork+0x2d/0x70 ret_from_fork_asm+0x1a/0x30 </TASK> So, for mm/vmstat:online, disable vmstat_work reliably on teardown and symmetrically enable it on startup. Link: https://lkml.kernel.org/r/20241221033321.4154409-1-koichiro.den@canonical.c… Signed-off-by: Koichiro Den <koichiro.den(a)canonical.com> Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- mm/vmstat.c | 3 ++- 1 file changed, 2 insertions(+), 1 deletion(-) --- a/mm/vmstat.c~vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep +++ a/mm/vmstat.c @@ -2148,13 +2148,14 @@ static int vmstat_cpu_online(unsigned in if (!node_state(cpu_to_node(cpu), N_CPU)) { node_set_state(cpu_to_node(cpu), N_CPU); } + enable_delayed_work(&per_cpu(vmstat_work, cpu)); return 0; } static int vmstat_cpu_down_prep(unsigned int cpu) { - cancel_delayed_work_sync(&per_cpu(vmstat_work, cpu)); + disable_delayed_work_sync(&per_cpu(vmstat_work, cpu)); return 0; } _ Patches currently in -mm which might be from koichiro.den(a)canonical.com are vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep.patch hugetlb-prioritize-surplus-allocation-from-current-node.patch

6 months, 1 week

1
0
0 0

+ vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep.patch added to mm-hotfixes-unstable branch

by Andrew Morton

The patch titled Subject: vmstat: disable vmstat_work on vmstat_cpu_down_prep() has been added to the -mm mm-hotfixes-unstable branch. Its filename is vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep.patch This patch will shortly appear at https://git.kernel.org/pub/scm/linux/kernel/git/akpm/25-new.git/tree/patche… This patch will later appear in the mm-hotfixes-unstable branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm Before you just go and hit "reply", please: a) Consider who else should be cc'ed b) Prefer to cc a suitable mailing list as well c) Ideally: find the original patch on the mailing list and do a reply-to-all to that, adding suitable additional cc's *** Remember to use Documentation/process/submit-checklist.rst when testing your code *** The -mm tree is included into linux-next via the mm-everything branch at git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm and is updated there every 2-3 working days ------------------------------------------------------ From: Koichiro Den <koichiro.den(a)canonical.com> Subject: vmstat: disable vmstat_work on vmstat_cpu_down_prep() Date: Fri, 20 Dec 2024 22:42:34 +0900 Even after mm/vmstat:online teardown, shepherd may still queue work for the dying cpu until the cpu is removed from online mask. While it's quite rare, this means that after unbind_workers() unbinds a per-cpu kworker, it potentially runs vmstat_update for the dying CPU on an irrelevant cpu before entering STARTING section. When CONFIG_DEBUG_PREEMPT=y, it results in the following error with the backtrace. BUG: using smp_processor_id() in preemptible [00000000] code: \ kworker/7:3/1702 caller is refresh_cpu_vm_stats+0x235/0x5f0 CPU: 0 UID: 0 PID: 1702 Comm: kworker/7:3 Tainted: G Tainted: [N]=TEST Workqueue: mm_percpu_wq vmstat_update Call Trace: <TASK> dump_stack_lvl+0x8d/0xb0 check_preemption_disabled+0xce/0xe0 refresh_cpu_vm_stats+0x235/0x5f0 vmstat_update+0x17/0xa0 process_one_work+0x869/0x1aa0 worker_thread+0x5e5/0x1100 kthread+0x29e/0x380 ret_from_fork+0x2d/0x70 ret_from_fork_asm+0x1a/0x30 </TASK> So, disable vmstat_work reliably on vmstat_cpu_down_prep(). Link: https://lkml.kernel.org/r/20241220134234.3809621-1-koichiro.den@canonical.c… Signed-off-by: Koichiro Den <koichiro.den(a)canonical.com> Cc: Sebastian Andrzej Siewior <bigeasy(a)linutronix.de> Cc: <stable(a)vger.kernel.org> Signed-off-by: Andrew Morton <akpm(a)linux-foundation.org> --- mm/vmstat.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/mm/vmstat.c~vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep +++ a/mm/vmstat.c @@ -2154,7 +2154,7 @@ static int vmstat_cpu_online(unsigned in static int vmstat_cpu_down_prep(unsigned int cpu) { - cancel_delayed_work_sync(&per_cpu(vmstat_work, cpu)); + disable_delayed_work_sync(&per_cpu(vmstat_work, cpu)); return 0; } _ Patches currently in -mm which might be from koichiro.den(a)canonical.com are vmstat-disable-vmstat_work-on-vmstat_cpu_down_prep.patch hugetlb-prioritize-surplus-allocation-from-current-node.patch

6 months, 1 week

1
0
0 0

[PATCH 24/28] drm/amd/display: Add check for granularity in dml ceil/floor helpers

by Roman.Li＠amd.com

From: Roman Li <Roman.Li(a)amd.com> [Why] Wrapper functions for dcn_bw_ceil2() and dcn_bw_floor2() should check for granularity is non zero to avoid assert and divide-by-zero error in dcn_bw_ functions. [How] Add check for granularity 0. Cc: Mario Limonciello <mario.limonciello(a)amd.com> Cc: stable(a)vger.kernel.org Reviewed-by: Alvin Lee <alvin.lee2(a)amd.com> Signed-off-by: Roman Li <Roman.Li(a)amd.com> Tested-by: Daniel Wheeler <daniel.wheeler(a)amd.com> --- drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h index 072bd0539605..6b2ab4ec2b5f 100644 --- a/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h +++ b/drivers/gpu/drm/amd/display/dc/dml/dml_inline_defs.h @@ -66,11 +66,15 @@ static inline double dml_max5(double a, double b, double c, double d, double e) static inline double dml_ceil(double a, double granularity) { + if (granularity == 0) + return 0; return (double) dcn_bw_ceil2(a, granularity); } static inline double dml_floor(double a, double granularity) { + if (granularity == 0) + return 0; return (double) dcn_bw_floor2(a, granularity); } @@ -114,11 +118,15 @@ static inline double dml_ceil_2(double f) static inline double dml_ceil_ex(double x, double granularity) { + if (granularity == 0) + return 0; return (double) dcn_bw_ceil2(x, granularity); } static inline double dml_floor_ex(double x, double granularity) { + if (granularity == 0) + return 0; return (double) dcn_bw_floor2(x, granularity); } -- 2.34.1

6 months, 1 week

1
0
0 0

[PATCH 2/2] usb: typec: tcpm: fix the sender response time issue

by joswang

From: Jos Wang <joswang(a)lenovo.com> According to the USB PD3 CTS specification, the requirements for tSenderResponse are different in PD2 and PD3 modes, see Table 19 Timing Table & Calculations. For PD2 mode, the tSenderResponse min 24ms and max 30ms; for PD3 mode, the tSenderResponse min 27ms and max 33ms. For the "TEST.PD.PROT.SRC.2 Get_Source_Cap No Request" test item, after receiving the Source_Capabilities Message sent by the UUT, the tester deliberately does not send a Request Message in order to force the SenderResponse timer on the Source UUT to timeout. The Tester checks that a Hard Reset is detected between tSenderResponse min and max，the delay is between the last bit of the GoodCRC Message EOP has been sent and the first bit of Hard Reset SOP has been received. The current code does not distinguish between PD2 and PD3 modes, and tSenderResponse defaults to 60ms. This will cause this test item and the following tests to fail: TEST.PD.PROT.SRC3.2 SenderResponseTimer Timeout TEST.PD.PROT.SNK.6 SenderResponseTimer Timeout Considering factors such as SOC performance, i2c rate, and the speed of PD chip sending data, "pd2-sender-response-time-ms" and "pd3-sender-response-time-ms" DT time properties are added to allow users to define platform timing. For values that have not been explicitly defined in DT using this property, a default value of 27ms for PD2 tSenderResponse and 30ms for PD3 tSenderResponse is set. Fixes: f0690a25a140 ("staging: typec: USB Type-C Port Manager (tcpm)") Cc: stable(a)vger.kernel.org Signed-off-by: Jos Wang <joswang(a)lenovo.com> --- drivers/usb/typec/tcpm/tcpm.c | 50 +++++++++++++++++++++++------------ include/linux/usb/pd.h | 3 ++- 2 files changed, 35 insertions(+), 18 deletions(-) diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c index 6021eeb903fe..3a159bfcf382 100644 --- a/drivers/usb/typec/tcpm/tcpm.c +++ b/drivers/usb/typec/tcpm/tcpm.c @@ -314,12 +314,16 @@ struct pd_data { * @sink_wait_cap_time: Deadline (in ms) for tTypeCSinkWaitCap timer * @ps_src_wait_off_time: Deadline (in ms) for tPSSourceOff timer * @cc_debounce_time: Deadline (in ms) for tCCDebounce timer + * @pd2_sender_response_time: Deadline (in ms) for pd20 tSenderResponse timer + * @pd3_sender_response_time: Deadline (in ms) for pd30 tSenderResponse timer */ struct pd_timings { u32 sink_wait_cap_time; u32 ps_src_off_time; u32 cc_debounce_time; u32 snk_bc12_cmpletion_time; + u32 pd2_sender_response_time; + u32 pd3_sender_response_time; }; struct tcpm_port { @@ -3776,7 +3780,9 @@ static bool tcpm_send_queued_message(struct tcpm_port *port) } else if (port->pwr_role == TYPEC_SOURCE) { tcpm_ams_finish(port); tcpm_set_state(port, HARD_RESET_SEND, - PD_T_SENDER_RESPONSE); + port->negotiated_rev >= PD_REV30 ? + port->timings.pd3_sender_response_time : + port->timings.pd2_sender_response_time); } else { tcpm_ams_finish(port); } @@ -4619,6 +4625,9 @@ static void run_state_machine(struct tcpm_port *port) enum typec_pwr_opmode opmode; unsigned int msecs; enum tcpm_state upcoming_state; + u32 sender_response_time = port->negotiated_rev >= PD_REV30 ? + port->timings.pd3_sender_response_time : + port->timings.pd2_sender_response_time; if (port->tcpc->check_contaminant && port->state != CHECK_CONTAMINANT) port->potential_contaminant = ((port->enter_state == SRC_ATTACH_WAIT && @@ -5113,7 +5122,7 @@ static void run_state_machine(struct tcpm_port *port) tcpm_set_state(port, SNK_WAIT_CAPABILITIES, 0); } else { tcpm_set_state_cond(port, hard_reset_state(port), - PD_T_SENDER_RESPONSE); + sender_response_time); } break; case SNK_NEGOTIATE_PPS_CAPABILITIES: @@ -5135,7 +5144,7 @@ static void run_state_machine(struct tcpm_port *port) tcpm_set_state(port, SNK_READY, 0); } else { tcpm_set_state_cond(port, hard_reset_state(port), - PD_T_SENDER_RESPONSE); + sender_response_time); } break; case SNK_TRANSITION_SINK: @@ -5387,7 +5396,7 @@ static void run_state_machine(struct tcpm_port *port) port->message_id_prime = 0; port->rx_msgid_prime = -1; tcpm_pd_send_control(port, PD_CTRL_SOFT_RESET, TCPC_TX_SOP_PRIME); - tcpm_set_state_cond(port, ready_state(port), PD_T_SENDER_RESPONSE); + tcpm_set_state_cond(port, ready_state(port), sender_response_time); } else { port->message_id = 0; port->rx_msgid = -1; @@ -5398,7 +5407,7 @@ static void run_state_machine(struct tcpm_port *port) tcpm_set_state_cond(port, hard_reset_state(port), 0); else tcpm_set_state_cond(port, hard_reset_state(port), - PD_T_SENDER_RESPONSE); + sender_response_time); } break; @@ -5409,8 +5418,7 @@ static void run_state_machine(struct tcpm_port *port) port->send_discover = true; port->send_discover_prime = false; } - tcpm_set_state_cond(port, DR_SWAP_SEND_TIMEOUT, - PD_T_SENDER_RESPONSE); + tcpm_set_state_cond(port, DR_SWAP_SEND_TIMEOUT, sender_response_time); break; case DR_SWAP_ACCEPT: tcpm_pd_send_control(port, PD_CTRL_ACCEPT, TCPC_TX_SOP); @@ -5444,7 +5452,7 @@ static void run_state_machine(struct tcpm_port *port) tcpm_set_state(port, ERROR_RECOVERY, 0); break; } - tcpm_set_state_cond(port, FR_SWAP_SEND_TIMEOUT, PD_T_SENDER_RESPONSE); + tcpm_set_state_cond(port, FR_SWAP_SEND_TIMEOUT, sender_response_time); break; case FR_SWAP_SEND_TIMEOUT: tcpm_set_state(port, ERROR_RECOVERY, 0); @@ -5475,8 +5483,7 @@ static void run_state_machine(struct tcpm_port *port) break; case PR_SWAP_SEND: tcpm_pd_send_control(port, PD_CTRL_PR_SWAP, TCPC_TX_SOP); - tcpm_set_state_cond(port, PR_SWAP_SEND_TIMEOUT, - PD_T_SENDER_RESPONSE); + tcpm_set_state_cond(port, PR_SWAP_SEND_TIMEOUT, sender_response_time); break; case PR_SWAP_SEND_TIMEOUT: tcpm_swap_complete(port, -ETIMEDOUT); @@ -5574,8 +5581,7 @@ static void run_state_machine(struct tcpm_port *port) break; case VCONN_SWAP_SEND: tcpm_pd_send_control(port, PD_CTRL_VCONN_SWAP, TCPC_TX_SOP); - tcpm_set_state(port, VCONN_SWAP_SEND_TIMEOUT, - PD_T_SENDER_RESPONSE); + tcpm_set_state(port, VCONN_SWAP_SEND_TIMEOUT, sender_response_time); break; case VCONN_SWAP_SEND_TIMEOUT: tcpm_swap_complete(port, -ETIMEDOUT); @@ -5656,23 +5662,21 @@ static void run_state_machine(struct tcpm_port *port) break; case GET_STATUS_SEND: tcpm_pd_send_control(port, PD_CTRL_GET_STATUS, TCPC_TX_SOP); - tcpm_set_state(port, GET_STATUS_SEND_TIMEOUT, - PD_T_SENDER_RESPONSE); + tcpm_set_state(port, GET_STATUS_SEND_TIMEOUT, sender_response_time); break; case GET_STATUS_SEND_TIMEOUT: tcpm_set_state(port, ready_state(port), 0); break; case GET_PPS_STATUS_SEND: tcpm_pd_send_control(port, PD_CTRL_GET_PPS_STATUS, TCPC_TX_SOP); - tcpm_set_state(port, GET_PPS_STATUS_SEND_TIMEOUT, - PD_T_SENDER_RESPONSE); + tcpm_set_state(port, GET_PPS_STATUS_SEND_TIMEOUT, sender_response_time); break; case GET_PPS_STATUS_SEND_TIMEOUT: tcpm_set_state(port, ready_state(port), 0); break; case GET_SINK_CAP: tcpm_pd_send_control(port, PD_CTRL_GET_SINK_CAP, TCPC_TX_SOP); - tcpm_set_state(port, GET_SINK_CAP_TIMEOUT, PD_T_SENDER_RESPONSE); + tcpm_set_state(port, GET_SINK_CAP_TIMEOUT, sender_response_time); break; case GET_SINK_CAP_TIMEOUT: port->sink_cap_done = true; @@ -7109,6 +7113,18 @@ static void tcpm_fw_get_timings(struct tcpm_port *port, struct fwnode_handle *fw ret = fwnode_property_read_u32(fwnode, "sink-bc12-completion-time-ms", &val); if (!ret) port->timings.snk_bc12_cmpletion_time = val; + + ret = fwnode_property_read_u32(fwnode, "pd2-sender-response-time-ms", &val); + if (!ret) + port->timings.pd2_sender_response_time = val; + else + port->timings.pd2_sender_response_time = PD_T_PD2_SENDER_RESPONSE; + + ret = fwnode_property_read_u32(fwnode, "pd3-sender-response-time-ms", &val); + if (!ret) + port->timings.pd3_sender_response_time = val; + else + port->timings.pd3_sender_response_time = PD_T_PD3_SENDER_RESPONSE; } static int tcpm_fw_get_caps(struct tcpm_port *port, struct fwnode_handle *fwnode) diff --git a/include/linux/usb/pd.h b/include/linux/usb/pd.h index d50098fb16b5..9c599e851b9a 100644 --- a/include/linux/usb/pd.h +++ b/include/linux/usb/pd.h @@ -457,7 +457,6 @@ static inline unsigned int rdo_max_power(u32 rdo) #define PD_T_NO_RESPONSE 5000 /* 4.5 - 5.5 seconds */ #define PD_T_DB_DETECT 10000 /* 10 - 15 seconds */ #define PD_T_SEND_SOURCE_CAP 150 /* 100 - 200 ms */ -#define PD_T_SENDER_RESPONSE 60 /* 24 - 30 ms, relaxed */ #define PD_T_RECEIVER_RESPONSE 15 /* 15ms max */ #define PD_T_SOURCE_ACTIVITY 45 #define PD_T_SINK_ACTIVITY 135 @@ -491,6 +490,8 @@ static inline unsigned int rdo_max_power(u32 rdo) #define PD_T_CC_DEBOUNCE 200 /* 100 - 200 ms */ #define PD_T_PD_DEBOUNCE 20 /* 10 - 20 ms */ #define PD_T_TRY_CC_DEBOUNCE 15 /* 10 - 20 ms */ +#define PD_T_PD2_SENDER_RESPONSE 27 /* PD20 spec 24 - 30 ms */ +#define PD_T_PD3_SENDER_RESPONSE 30 /* PD30 spec 27 - 33 ms */ #define PD_N_CAPS_COUNT (PD_T_NO_RESPONSE / PD_T_SEND_SOURCE_CAP) #define PD_N_HARD_RESET_COUNT 2 -- 2.17.1

6 months, 1 week

4
3
0 0

[PATCH v3 00/28] usb: gadget: f_tcm: Enhance UASP driver

by Thinh Nguyen

Apologies for the delay; after two years and multiple requests to resume this series, I squeezed some time to push an update. This series applies on top of Greg's usb-testing branch. If possible, please help test this series and get this merged as my resources are nil for this work. Example Bringup Steps ===================== To test UASP, here's an example perl script snippet to bring it up. Note: the script was cut down and quickly rewritten, so sorry if I make mistakes. my $MY_UAS_VID = xxxx; my $MY_UAS_PID = yyyy; my $SERIAL = "1234"; my $VENDOR = "VENDOR"; my $MY_VER = "VER"; my $vendor_id = "my_vid"; my $product_id = "my_pid"; my $revision = "my_rev"; # Must update: my $backing_storage = "/tmp/some_file"; my $backing_storage_size = 1024*1024*16; my $use_ramdisk = 0; my $g = "/sys/kernel/config/usb_gadget/g1"; system("modprobe libcomposite"); system("modprobe usb_f_tcm"); system("mkdir -p $g"); system("mkdir -p $g/configs/c.1"); system("mkdir -p $g/functions/tcm.0"); system("mkdir -p $g/strings/0x409"); system("mkdir -p $g/configs/c.1/strings/0x409"); my $tp = "/sys/kernel/config/target/usb_gadget/naa.0/tpgt_1"; my $tf; my $ctrl; if ($use_ramdisk) { $tf = "/sys/kernel/config/target/core/rd_mcp_0/ramdisk"; $ctrl = 'rd_pages=524288'; } else { $tf = "/sys/kernel/config/target/core/fileio_0/fileio"; $ctrl = 'fd_dev_name=$backing_storage,fd_dev_size=$backing_storage_size,fd_async_io=1'; } system("mkdir -p /etc/target"); system("mkdir -p $tp"); system("mkdir -p $tf"); system("mkdir -p $tp/lun/lun_0"); system("echo naa.0 > $tp/nexus"); system("echo $ctrl > $tf/control"); system("echo 1 > $tf/attrib/emulate_ua_intlck_ctrl"); system("echo 123 > $tf/wwn/vpd_unit_serial"); system("echo $vendor_id > $tf/wwn/vendor_id"); system("echo $product_id > $tf/wwn/product_id"); system("echo $revision > $tf/wwn/revision"); system("echo 1 > $tf/enable"); system("ln -s $tf $tp/lun/lun_0/virtual_scsi_port"); system("echo 1 > $tp/enable"); system("echo $MY_UAS_PID > $g/idProduct"); system("ln -s $g/functions/tcm.0 $g/configs/c.1"); system("echo $MY_UAS_VID > $g/idVendor"); system("echo $SERIAL > $g/strings/0x409/serialnumber"); system("echo $VENDOR > $g/strings/0x409/manufacturer"); system("echo \"$MY_VER\" > $g/strings/0x409/product"); system("echo \"Conf 1\" > $g/configs/c.1/strings/0x409/configuration"); system("echo super-speed-plus > $g/max_speed"); # Make sure the UDC is available system("echo $my_udc > $g/UDC"); Target Subsystem Fixes ====================== I have eliminated unnecessary changes related to the Target subsystem and reworked f_tcm to minimize the modifications required in the Target subsystem. There are unimplemented Task Management Requests in the Target subsystem, but the basic flow should still work. Regardless, you should still need to apply at least these 2 fixes: 1) Fix Data Corruption ---------------------- Properly increment the "len" base on the command requested length instead of the SG entry length. If you're using File backend, then you need to fix target_core_file. If you're using other backend such as Ramdisk, then you need a similar fix there. --- drivers/target/target_core_file.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/target/target_core_file.c b/drivers/target/target_core_file.c index 2d78ef74633c..d9fc048c1734 100644 --- a/drivers/target/target_core_file.c +++ b/drivers/target/target_core_file.c @@ -283,7 +283,12 @@ fd_execute_rw_aio(struct se_cmd *cmd, struct scatterlist *sgl, u32 sgl_nents, for_each_sg(sgl, sg, sgl_nents, i) { bvec_set_page(&aio_cmd->bvecs[i], sg_page(sg), sg->length, sg->offset); - len += sg->length; + if (len + sg->length >= cmd->data_length) { + len = cmd->data_length; + break; + } else { + len += sg->length; + } } iov_iter_bvec(&iter, is_write, aio_cmd->bvecs, sgl_nents, len); @@ -328,7 +333,12 @@ static int fd_do_rw(struct se_cmd *cmd, struct file *fd, for_each_sg(sgl, sg, sgl_nents, i) { bvec_set_page(&bvec[i], sg_page(sg), sg->length, sg->offset); - len += sg->length; + if (len + sg->length >= data_length) { + len = data_length; + break; + } else { + len += sg->length; + } } iov_iter_bvec(&iter, is_write, bvec, sgl_nents, len); -- 2) Fix Sense Data Length ------------------------ The transport_get_sense_buffer() and transport_copy_sense_to_cmd() take sense data length to be the allocated sense buffer length TRANSPORT_SENSE_BUFFER. However, the sense data length is depending on the sense data description. Check the sense data to set the proper cmd->scsi_sense_length. See SPC4-r37 section 4.5.2.1. --- drivers/target/target_core_transport.c | 19 +++++++++++++++---- 1 file changed, 15 insertions(+), 4 deletions(-) diff --git a/drivers/target/target_core_transport.c b/drivers/target/target_core_transport.c index 8d8f4ad4f59e..da75d6873ab5 100644 --- a/drivers/target/target_core_transport.c +++ b/drivers/target/target_core_transport.c @@ -804,8 +804,6 @@ static unsigned char *transport_get_sense_buffer(struct se_cmd *cmd) if (cmd->se_cmd_flags & SCF_SENT_CHECK_CONDITION) return NULL; - cmd->scsi_sense_length = TRANSPORT_SENSE_BUFFER; - pr_debug("HBA_[%u]_PLUG[%s]: Requesting sense for SAM STATUS: 0x%02x\n", dev->se_hba->hba_id, dev->transport->name, cmd->scsi_status); return cmd->sense_buffer; @@ -824,7 +822,13 @@ void transport_copy_sense_to_cmd(struct se_cmd *cmd, unsigned char *sense) } cmd->se_cmd_flags |= SCF_TRANSPORT_TASK_SENSE; + + /* Sense data length = min sense data + additional sense data length */ + cmd->scsi_sense_length = min_t(u16, cmd_sense_buf[7] + 8, + TRANSPORT_SENSE_BUFFER); + memcpy(cmd_sense_buf, sense, cmd->scsi_sense_length); + spin_unlock_irqrestore(&cmd->t_state_lock, flags); } EXPORT_SYMBOL(transport_copy_sense_to_cmd); @@ -3521,12 +3525,19 @@ static void translate_sense_reason(struct se_cmd *cmd, sense_reason_t reason) cmd->se_cmd_flags |= SCF_EMULATED_TASK_SENSE; cmd->scsi_status = SAM_STAT_CHECK_CONDITION; - cmd->scsi_sense_length = TRANSPORT_SENSE_BUFFER; + scsi_build_sense_buffer(desc_format, buffer, key, asc, ascq); if (sd->add_sense_info) WARN_ON_ONCE(scsi_set_sense_information(buffer, - cmd->scsi_sense_length, + TRANSPORT_SENSE_BUFFER, cmd->sense_info) < 0); + /* + * CHECK CONDITION returns sense data, and sense data is minimum 8 + * bytes long plus additional Sense Data Length. + * See SPC4-r37 section 4.5.2.1. + */ + cmd->scsi_sense_length = min_t(u16, buffer[7] + 8, + TRANSPORT_SENSE_BUFFER); } int -- Changes in v3: - v2: https://lore.kernel.org/linux-usb/cover.1658192351.git.Thinh.Nguyen@synopsy… - Moved patches around so fixes patches go first - Use hashtable to map tag to uas stream - Move target_execute_cmd() out of interrupt context - Various cleanup - Additional fixes over the 2 years Thinh Nguyen (28): usb: gadget: f_tcm: Don't free command immediately usb: gadget: f_tcm: Translate error to sense usb: gadget: f_tcm: Decrement command ref count on cleanup usb: gadget: f_tcm: Fix Get/SetInterface return value usb: gadget: f_tcm: ep_autoconfig with fullspeed endpoint usb: gadget: f_tcm: Don't prepare BOT write request twice usb: gadget: f_tcm: Increase stream count usb: gadget: f_tcm: Increase bMaxBurst usb: gadget: f_tcm: Limit number of sessions usb: gadget: f_tcm: Get stream by sbitmap number usb: gadget: f_tcm: Don't set static stream_id usb: gadget: f_tcm: Allocate matching number of commands to streams usb: gadget: f_tcm: Handle multiple commands in parallel usb: gadget: f_tcm: Use extra number of commands usb: gadget: f_tcm: Return ATA cmd direction usb: gadget: f_tcm: Execute command on write completion usb: gadget: f_tcm: Minor cleanup redundant code usb: gadget: f_tcm: Handle abort command usb: gadget: f_tcm: Cleanup requests on ep disable usb: gadget: f_tcm: Stop proceeding further on -ESHUTDOWN usb: gadget: f_tcm: Save CPU ID per command usb: gadget: f_tcm: Send sense on cancelled transfer usb: gadget: f_tcm: Handle TASK_MANAGEMENT commands usb: gadget: f_tcm: Check overlapped command usb: gadget: f_tcm: Stall on invalid CBW usb: gadget: f_tcm: Requeue command request on error usb: gadget: f_tcm: Track BOT command kref usb: gadget: f_tcm: Refactor goto check_condition drivers/usb/gadget/function/f_tcm.c | 711 ++++++++++++++++++++-------- drivers/usb/gadget/function/tcm.h | 28 +- 2 files changed, 547 insertions(+), 192 deletions(-) base-commit: d8d936c51388442f769a81e512b505dcf87c6a51 -- 2.28.0

6 months, 1 week

3
10
0 0

[PATCH AUTOSEL 5.4 1/6] wifi: mac80211: wake the queues in case of failure in resume

by Sasha Levin

From: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> [ Upstream commit 220bf000530f9b1114fa2a1022a871c7ce8a0b38 ] In case we fail to resume, we'll WARN with "Hardware became unavailable during restart." and we'll wait until user space does something. It'll typically bring the interface down and up to recover. This won't work though because the queues are still stopped on IEEE80211_QUEUE_STOP_REASON_SUSPEND reason. Make sure we clear that reason so that we give a chance to the recovery to succeed. Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219447 Signed-off-by: Miri Korenblit <miriam.rachel.korenblit(a)intel.com> Link: https://patch.msgid.link/20241119173108.cd628f560f97.I76a15fdb92de450e53299… Signed-off-by: Johannes Berg <johannes.berg(a)intel.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- net/mac80211/util.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 63b66fd0a1ce..515fe1d539b4 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -2209,6 +2209,9 @@ int ieee80211_reconfig(struct ieee80211_local *local) WARN(1, "Hardware became unavailable upon resume. This could be a software issue prior to suspend or a hardware issue.\n"); else WARN(1, "Hardware became unavailable during restart.\n"); + ieee80211_wake_queues_by_reason(hw, IEEE80211_MAX_QUEUE_MAP, + IEEE80211_QUEUE_STOP_REASON_SUSPEND, + false); ieee80211_handle_reconfig_failure(local); return res; } -- 2.39.5

6 months, 1 week

1
5
0 0

[PATCH AUTOSEL 5.10 1/7] wifi: mac80211: wake the queues in case of failure in resume

by Sasha Levin

From: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> [ Upstream commit 220bf000530f9b1114fa2a1022a871c7ce8a0b38 ] In case we fail to resume, we'll WARN with "Hardware became unavailable during restart." and we'll wait until user space does something. It'll typically bring the interface down and up to recover. This won't work though because the queues are still stopped on IEEE80211_QUEUE_STOP_REASON_SUSPEND reason. Make sure we clear that reason so that we give a chance to the recovery to succeed. Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219447 Signed-off-by: Miri Korenblit <miriam.rachel.korenblit(a)intel.com> Link: https://patch.msgid.link/20241119173108.cd628f560f97.I76a15fdb92de450e53299… Signed-off-by: Johannes Berg <johannes.berg(a)intel.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- net/mac80211/util.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mac80211/util.c b/net/mac80211/util.c index e49355cbb1ce..0da845d9d486 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -2351,6 +2351,9 @@ int ieee80211_reconfig(struct ieee80211_local *local) WARN(1, "Hardware became unavailable upon resume. This could be a software issue prior to suspend or a hardware issue.\n"); else WARN(1, "Hardware became unavailable during restart.\n"); + ieee80211_wake_queues_by_reason(hw, IEEE80211_MAX_QUEUE_MAP, + IEEE80211_QUEUE_STOP_REASON_SUSPEND, + false); ieee80211_handle_reconfig_failure(local); return res; } -- 2.39.5

6 months, 1 week

1
6
0 0

[PATCH AUTOSEL 5.15 1/9] wifi: mac80211: wake the queues in case of failure in resume

by Sasha Levin

From: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> [ Upstream commit 220bf000530f9b1114fa2a1022a871c7ce8a0b38 ] In case we fail to resume, we'll WARN with "Hardware became unavailable during restart." and we'll wait until user space does something. It'll typically bring the interface down and up to recover. This won't work though because the queues are still stopped on IEEE80211_QUEUE_STOP_REASON_SUSPEND reason. Make sure we clear that reason so that we give a chance to the recovery to succeed. Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219447 Signed-off-by: Miri Korenblit <miriam.rachel.korenblit(a)intel.com> Link: https://patch.msgid.link/20241119173108.cd628f560f97.I76a15fdb92de450e53299… Signed-off-by: Johannes Berg <johannes.berg(a)intel.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- net/mac80211/util.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 85d3d2034d43..cc78d3cba45e 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -2374,6 +2374,9 @@ int ieee80211_reconfig(struct ieee80211_local *local) WARN(1, "Hardware became unavailable upon resume. This could be a software issue prior to suspend or a hardware issue.\n"); else WARN(1, "Hardware became unavailable during restart.\n"); + ieee80211_wake_queues_by_reason(hw, IEEE80211_MAX_QUEUE_MAP, + IEEE80211_QUEUE_STOP_REASON_SUSPEND, + false); ieee80211_handle_reconfig_failure(local); return res; } -- 2.39.5

6 months, 1 week

1
8
0 0

[PATCH AUTOSEL 6.1 01/12] wifi: mac80211: wake the queues in case of failure in resume

by Sasha Levin

From: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> [ Upstream commit 220bf000530f9b1114fa2a1022a871c7ce8a0b38 ] In case we fail to resume, we'll WARN with "Hardware became unavailable during restart." and we'll wait until user space does something. It'll typically bring the interface down and up to recover. This won't work though because the queues are still stopped on IEEE80211_QUEUE_STOP_REASON_SUSPEND reason. Make sure we clear that reason so that we give a chance to the recovery to succeed. Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach(a)intel.com> Closes: https://bugzilla.kernel.org/show_bug.cgi?id=219447 Signed-off-by: Miri Korenblit <miriam.rachel.korenblit(a)intel.com> Link: https://patch.msgid.link/20241119173108.cd628f560f97.I76a15fdb92de450e53299… Signed-off-by: Johannes Berg <johannes.berg(a)intel.com> Signed-off-by: Sasha Levin <sashal(a)kernel.org> --- net/mac80211/util.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/net/mac80211/util.c b/net/mac80211/util.c index 738f1f139a90..e8326e09d1b3 100644 --- a/net/mac80211/util.c +++ b/net/mac80211/util.c @@ -2436,6 +2436,9 @@ int ieee80211_reconfig(struct ieee80211_local *local) WARN(1, "Hardware became unavailable upon resume. This could be a software issue prior to suspend or a hardware issue.\n"); else WARN(1, "Hardware became unavailable during restart.\n"); + ieee80211_wake_queues_by_reason(hw, IEEE80211_MAX_QUEUE_MAP, + IEEE80211_QUEUE_STOP_REASON_SUSPEND, + false); ieee80211_handle_reconfig_failure(local); return res; } -- 2.39.5

6 months, 1 week

1
11
0 0

2025

2024

2023

2022

2021

2020

2019

2018

2017

Linux-stable-mirror December 2024