dpc.c 9.1 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
2 3
/*
 * PCI Express Downstream Port Containment services driver
4 5
 * Author: Keith Busch <keith.busch@intel.com>
 *
6 7 8 9 10
 * Copyright (C) 2016 Intel Corp.
 */

#include <linux/delay.h>
#include <linux/interrupt.h>
11
#include <linux/init.h>
12
#include <linux/pci.h>
13

14
#include "portdrv.h"
K
Keith Busch 已提交
15
#include "../pci.h"
16 17 18

struct dpc_dev {
	struct pcie_device	*dev;
19
	struct work_struct	work;
20
	u16			cap_pos;
21
	bool			rp_extensions;
D
Dongdong Liu 已提交
22
	u32			rp_pio_status;
23
	u8			rp_log_size;
D
Dongdong Liu 已提交
24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45
};

static const char * const rp_pio_error_string[] = {
	"Configuration Request received UR Completion",	 /* Bit Position 0  */
	"Configuration Request received CA Completion",	 /* Bit Position 1  */
	"Configuration Request Completion Timeout",	 /* Bit Position 2  */
	NULL,
	NULL,
	NULL,
	NULL,
	NULL,
	"I/O Request received UR Completion",		 /* Bit Position 8  */
	"I/O Request received CA Completion",		 /* Bit Position 9  */
	"I/O Request Completion Timeout",		 /* Bit Position 10 */
	NULL,
	NULL,
	NULL,
	NULL,
	NULL,
	"Memory Request received UR Completion",	 /* Bit Position 16 */
	"Memory Request received CA Completion",	 /* Bit Position 17 */
	"Memory Request Completion Timeout",		 /* Bit Position 18 */
46 47
};

48 49 50 51
static int dpc_wait_rp_inactive(struct dpc_dev *dpc)
{
	unsigned long timeout = jiffies + HZ;
	struct pci_dev *pdev = dpc->dev->port;
52
	struct device *dev = &dpc->dev->device;
53
	u16 cap = dpc->cap_pos, status;
54

55
	pci_read_config_word(pdev, cap + PCI_EXP_DPC_STATUS, &status);
56 57 58
	while (status & PCI_EXP_DPC_RP_BUSY &&
					!time_after(jiffies, timeout)) {
		msleep(10);
59
		pci_read_config_word(pdev, cap + PCI_EXP_DPC_STATUS, &status);
60 61
	}
	if (status & PCI_EXP_DPC_RP_BUSY) {
62
		dev_warn(dev, "DPC root port still busy\n");
63 64 65 66 67
		return -EBUSY;
	}
	return 0;
}

68
static void dpc_wait_link_inactive(struct dpc_dev *dpc)
69
{
70
	struct pci_dev *pdev = dpc->dev->port;
71

72
	pcie_wait_for_link(pdev, false);
73 74
}

75
static pci_ers_result_t dpc_reset_link(struct pci_dev *pdev)
76
{
77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
	struct dpc_dev *dpc;
	struct pcie_device *pciedev;
	struct device *devdpc;
	u16 cap, ctl;

	/*
	 * DPC disables the Link automatically in hardware, so it has
	 * already been reset by the time we get here.
	 */
	devdpc = pcie_port_find_device(pdev, PCIE_PORT_SERVICE_DPC);
	pciedev = to_pcie_device(devdpc);
	dpc = get_service_data(pciedev);
	cap = dpc->cap_pos;

	/*
	 * Wait until the Link is inactive, then clear DPC Trigger Status
	 * to allow the Port to leave DPC.
	 */
95
	dpc_wait_link_inactive(dpc);
96

97
	if (dpc->rp_extensions && dpc_wait_rp_inactive(dpc))
98
		return PCI_ERS_RESULT_DISCONNECT;
99
	if (dpc->rp_extensions && dpc->rp_pio_status) {
100 101
		pci_write_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_STATUS,
				       dpc->rp_pio_status);
D
Dongdong Liu 已提交
102 103 104
		dpc->rp_pio_status = 0;
	}

105
	pci_write_config_word(pdev, cap + PCI_EXP_DPC_STATUS,
106
			      PCI_EXP_DPC_STATUS_TRIGGER);
107

108 109
	pci_read_config_word(pdev, cap + PCI_EXP_DPC_CTL, &ctl);
	pci_write_config_word(pdev, cap + PCI_EXP_DPC_CTL,
110
			      ctl | PCI_EXP_DPC_CTL_INT_EN);
111 112 113 114 115 116 117 118 119 120 121

	return PCI_ERS_RESULT_RECOVERED;
}

static void dpc_work(struct work_struct *work)
{
	struct dpc_dev *dpc = container_of(work, struct dpc_dev, work);
	struct pci_dev *pdev = dpc->dev->port;

	/* We configure DPC so it only triggers on ERR_FATAL */
	pcie_do_fatal_recovery(pdev, PCIE_PORT_SERVICE_DPC);
122 123
}

124
static void dpc_process_rp_pio_error(struct dpc_dev *dpc)
D
Dongdong Liu 已提交
125
{
126
	struct device *dev = &dpc->dev->device;
D
Dongdong Liu 已提交
127
	struct pci_dev *pdev = dpc->dev->port;
128 129
	u16 cap = dpc->cap_pos, dpc_status, first_error;
	u32 status, mask, sev, syserr, exc, dw0, dw1, dw2, dw3, log, prefix;
D
Dongdong Liu 已提交
130 131
	int i;

132 133
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_STATUS, &status);
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_MASK, &mask);
134
	dev_err(dev, "rp_pio_status: %#010x, rp_pio_mask: %#010x\n",
135
		status, mask);
D
Dongdong Liu 已提交
136

137
	dpc->rp_pio_status = status;
138

139 140 141
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_SEVERITY, &sev);
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_SYSERROR, &syserr);
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_EXCEPTION, &exc);
142
	dev_err(dev, "RP PIO severity=%#010x, syserror=%#010x, exception=%#010x\n",
143
		sev, syserr, exc);
D
Dongdong Liu 已提交
144 145

	/* Get First Error Pointer */
146
	pci_read_config_word(pdev, cap + PCI_EXP_DPC_STATUS, &dpc_status);
147
	first_error = (dpc_status & 0x1f00) >> 8;
D
Dongdong Liu 已提交
148

149
	status &= ~mask;
150 151 152
	for (i = 0; i < ARRAY_SIZE(rp_pio_error_string); i++) {
		if (status & (1 << i))
			dev_err(dev, "[%2d] %s%s\n", i, rp_pio_error_string[i],
153
				first_error == i ? " (First)" : "");
154 155
	}

156
	if (dpc->rp_log_size < 4)
D
Dongdong Liu 已提交
157
		return;
158
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_HEADER_LOG,
159
			      &dw0);
160
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_HEADER_LOG + 4,
161
			      &dw1);
162
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_HEADER_LOG + 8,
163
			      &dw2);
164
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_HEADER_LOG + 12,
165
			      &dw3);
166
	dev_err(dev, "TLP Header: %#010x %#010x %#010x %#010x\n",
167
		dw0, dw1, dw2, dw3);
D
Dongdong Liu 已提交
168

169 170
	if (dpc->rp_log_size < 5)
		return;
171 172
	pci_read_config_dword(pdev, cap + PCI_EXP_DPC_RP_PIO_IMPSPEC_LOG, &log);
	dev_err(dev, "RP PIO ImpSpec Log %#010x\n", log);
173

174
	for (i = 0; i < dpc->rp_log_size - 5; i++) {
D
Dongdong Liu 已提交
175
		pci_read_config_dword(pdev,
176 177
			cap + PCI_EXP_DPC_RP_PIO_TLPPREFIX_LOG, &prefix);
		dev_err(dev, "TLP Prefix Header: dw%d, %#010x\n", i, prefix);
178
	}
D
Dongdong Liu 已提交
179 180
}

181 182 183 184
static irqreturn_t dpc_irq(int irq, void *context)
{
	struct dpc_dev *dpc = (struct dpc_dev *)context;
	struct pci_dev *pdev = dpc->dev->port;
185
	struct device *dev = &dpc->dev->device;
186
	u16 cap = dpc->cap_pos, ctl, status, source, reason, ext_reason;
187

188
	pci_read_config_word(pdev, cap + PCI_EXP_DPC_CTL, &ctl);
189 190 191

	if (!(ctl & PCI_EXP_DPC_CTL_INT_EN) || ctl == (u16)(~0))
		return IRQ_NONE;
192

193
	pci_read_config_word(pdev, cap + PCI_EXP_DPC_STATUS, &status);
194 195 196 197 198

	if (!(status & PCI_EXP_DPC_STATUS_INTERRUPT))
		return IRQ_NONE;

	if (!(status & PCI_EXP_DPC_STATUS_TRIGGER)) {
199
		pci_write_config_word(pdev, cap + PCI_EXP_DPC_STATUS,
200 201 202 203
				      PCI_EXP_DPC_STATUS_INTERRUPT);
		return IRQ_HANDLED;
	}

204
	pci_write_config_word(pdev, cap + PCI_EXP_DPC_CTL,
205 206
			      ctl & ~PCI_EXP_DPC_CTL_INT_EN);

207
	pci_read_config_word(pdev, cap + PCI_EXP_DPC_SOURCE_ID,
208 209
			     &source);

210
	dev_info(dev, "DPC containment event, status:%#06x source:%#06x\n",
211 212
		status, source);

213 214
	reason = (status & PCI_EXP_DPC_STATUS_TRIGGER_RSN) >> 1;
	ext_reason = (status & PCI_EXP_DPC_STATUS_TRIGGER_RSN_EXT) >> 5;
215 216 217 218 219 220 221 222 223

	dev_warn(dev, "DPC %s detected, remove downstream devices\n",
		 (reason == 0) ? "unmasked uncorrectable error" :
		 (reason == 1) ? "ERR_NONFATAL" :
		 (reason == 2) ? "ERR_FATAL" :
		 (ext_reason == 0) ? "RP PIO error" :
		 (ext_reason == 1) ? "software trigger" :
				     "reserved error");
	/* show RP PIO error detail information */
224
	if (dpc->rp_extensions && reason == 3 && ext_reason == 0)
225 226
		dpc_process_rp_pio_error(dpc);

227 228 229
	pci_write_config_word(pdev, cap + PCI_EXP_DPC_STATUS,
			      PCI_EXP_DPC_STATUS_INTERRUPT);

230 231
	schedule_work(&dpc->work);

232 233 234 235 236 237 238 239
	return IRQ_HANDLED;
}

#define FLAG(x, y) (((x) & (y)) ? '+' : '-')
static int dpc_probe(struct pcie_device *dev)
{
	struct dpc_dev *dpc;
	struct pci_dev *pdev = dev->port;
240
	struct device *device = &dev->device;
241 242 243
	int status;
	u16 ctl, cap;

244 245 246
	if (pcie_aer_get_firmware_first(pdev))
		return -ENOTSUPP;

247
	dpc = devm_kzalloc(device, sizeof(*dpc), GFP_KERNEL);
248 249 250 251 252
	if (!dpc)
		return -ENOMEM;

	dpc->cap_pos = pci_find_ext_capability(pdev, PCI_EXT_CAP_ID_DPC);
	dpc->dev = dev;
253
	INIT_WORK(&dpc->work, dpc_work);
254 255
	set_service_data(dev, dpc);

256
	status = devm_request_irq(device, dev->irq, dpc_irq, IRQF_SHARED,
257
				  "pcie-dpc", dpc);
258
	if (status) {
259
		dev_warn(device, "request IRQ%d failed: %d\n", dev->irq,
260
			 status);
261
		return status;
262 263 264 265 266
	}

	pci_read_config_word(pdev, dpc->cap_pos + PCI_EXP_DPC_CAP, &cap);
	pci_read_config_word(pdev, dpc->cap_pos + PCI_EXP_DPC_CTL, &ctl);

267
	dpc->rp_extensions = (cap & PCI_EXP_DPC_CAP_RP_EXT);
268 269 270 271 272 273 274 275
	if (dpc->rp_extensions) {
		dpc->rp_log_size = (cap & PCI_EXP_DPC_RP_PIO_LOG_SIZE) >> 8;
		if (dpc->rp_log_size < 4 || dpc->rp_log_size > 9) {
			dev_err(device, "RP PIO log size %u is invalid\n",
				dpc->rp_log_size);
			dpc->rp_log_size = 0;
		}
	}
276

277
	ctl = (ctl & 0xfff4) | PCI_EXP_DPC_CTL_EN_FATAL | PCI_EXP_DPC_CTL_INT_EN;
278 279
	pci_write_config_word(pdev, dpc->cap_pos + PCI_EXP_DPC_CTL, ctl);

280
	dev_info(device, "DPC error containment capabilities: Int Msg #%d, RPExt%c PoisonedTLP%c SwTrigger%c RP PIO Log %d, DL_ActiveErr%c\n",
281
		cap & PCI_EXP_DPC_IRQ, FLAG(cap, PCI_EXP_DPC_CAP_RP_EXT),
282
		FLAG(cap, PCI_EXP_DPC_CAP_POISONED_TLP),
283
		FLAG(cap, PCI_EXP_DPC_CAP_SW_TRIGGER), dpc->rp_log_size,
284 285 286 287 288 289 290 291 292 293 294
		FLAG(cap, PCI_EXP_DPC_CAP_DL_ACTIVE));
	return status;
}

static void dpc_remove(struct pcie_device *dev)
{
	struct dpc_dev *dpc = get_service_data(dev);
	struct pci_dev *pdev = dev->port;
	u16 ctl;

	pci_read_config_word(pdev, dpc->cap_pos + PCI_EXP_DPC_CTL, &ctl);
295
	ctl &= ~(PCI_EXP_DPC_CTL_EN_FATAL | PCI_EXP_DPC_CTL_INT_EN);
296 297 298 299 300
	pci_write_config_word(pdev, dpc->cap_pos + PCI_EXP_DPC_CTL, ctl);
}

static struct pcie_port_service_driver dpcdriver = {
	.name		= "dpc",
301
	.port_type	= PCIE_ANY_PORT,
302 303 304
	.service	= PCIE_PORT_SERVICE_DPC,
	.probe		= dpc_probe,
	.remove		= dpc_remove,
305
	.reset_link	= dpc_reset_link,
306 307 308 309 310 311
};

static int __init dpc_service_init(void)
{
	return pcie_port_service_register(&dpcdriver);
}
312
device_initcall(dpc_service_init);