flashrom March 2017

flashrom@flashrom.org

21 participants
36 discussions

[PATCH v2] nicintel_eeprom: Support for I210 emulated EEprom

by Ricardo Ribalda Delgado

On the I210 family there is no MAC EEprom, instead there is a big flash (typically around 16Mb) with contents of the old MAC plus other stuff. There is an interface to program the whole flash, but once it is programmed it enters a "Secure Mode" that disables the interface. Luckily, the section with the MAC can still be updated via the EEprom interface. This patch adds support for this interface. root@qt5022-fglrx:~# ./flashrom -p nicintel_eeprom:pci=01:0.0 -w kk.raw -V flashrom v0.9.9-unknown on Linux 4.10.0-qtec-standard (x86_64) flashrom is free software, get the source code at https://flashrom.org flashrom was built with libpci 3.4.1, GCC 5.3.0, little endian Command line (5 args): ./flashrom -p nicintel_eeprom:pci=01:0.0 -w kk.raw -V Calibrating delay loop... OS timer resolution is 1 usecs, 1856M loops per second, 10 myus = 10 us, 100 myus = 102 us, 1000 myus = 1017 us, 10000 myus = 10044 us, 4 myus = 4 us, OK. Initializing nicintel_eeprom programmer Found "Intel I210 Gigabit Network Connection" (8086:1533, BDF 01:00.0). Requested BAR is of type MEM, 32bit, not prefetchable Requested BAR is of type MEM, 32bit, not prefetchable The following protocols are supported: Programmer-specific. Probing for Programmer Opaque flash chip, 0 kB: Found Programmer flash chip "Opaque flash chip" (4 kB, Programmer-specific) on nicintel_eeprom. Found Programmer flash chip "Opaque flash chip" (4 kB, Programmer-specific). Reading old flash chip contents... done. Erasing and writing flash chip... Trying erase function 0... 0x000000-0x000fff:W Erase/write done. Verifying flash... VERIFIED. Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda(a)gmail.com> --- v2: Please completely ignore v1. nicintel_eeprom.c | 174 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 168 insertions(+), 6 deletions(-) diff --git a/nicintel_eeprom.c b/nicintel_eeprom.c index b5d4202b01e0..a63e937f1ce4 100644 --- a/nicintel_eeprom.c +++ b/nicintel_eeprom.c @@ -25,6 +25,12 @@ * 4.7: Access to shared resources (FIXME: we should probably use this semaphore interface) * 7.4: Register Descriptions */ +/* + * Datasheet: Intel Ethernet Controller I210: Datasheet + * 8.4.3: EEPROM-Mode Read Register + * 8.4.6: EEPROM-Mode Write Register + * Write process inspired on kernel e1000_i210.c + */ #include <stdlib.h> #include <unistd.h> @@ -34,10 +40,11 @@ #include "hwaccess.h" #define PCI_VENDOR_ID_INTEL 0x8086 -#define MEMMAP_SIZE (0x14 + 3) /* Only EEC and EERD are needed. */ +#define MEMMAP_SIZE 0x1c /* Only EEC, EERD and EEWR are needed. */ #define EEC 0x10 /* EEPROM/Flash Control Register */ #define EERD 0x14 /* EEPROM Read Register */ +#define EEWR 0x18 /* EEPROM Write Register */ /* EPROM/Flash Control Register bits */ #define EE_SCK 0 @@ -49,6 +56,8 @@ #define EE_PRES 8 #define EE_SIZE 11 #define EE_SIZE_MASK 0xf +#define EE_FLUPD 23 +#define EE_FLUDONE 26 /* EEPROM Read Register bits */ #define EERD_START 0 @@ -56,11 +65,18 @@ #define EERD_ADDR 2 #define EERD_DATA 16 +/* EEPROM Write Register bits */ +#define EEWR_CMDV 0 +#define EEWR_DONE 1 +#define EEWR_ADDR 2 +#define EEWR_DATA 16 + #define BIT(x) (1<<x) #define EE_PAGE_MASK 0x3f static uint8_t *nicintel_eebar; static struct pci_dev *nicintel_pci; +static bool done_i20_write = false; #define UNPROG_DEVICE 0x1509 @@ -71,10 +87,35 @@ const struct dev_entry nics_intel_ee[] = { {PCI_VENDOR_ID_INTEL, 0x1511, NT , "Intel", "82580 Quad Gigabit Ethernet Controller (Ext. PHY)"}, {PCI_VENDOR_ID_INTEL, 0x1511, NT , "Intel", "82580 Dual Gigabit Ethernet Controller (Copper)"}, {PCI_VENDOR_ID_INTEL, UNPROG_DEVICE, OK, "Intel", "Unprogrammed 82580 Quad/Dual Gigabit Ethernet Controller"}, + {PCI_VENDOR_ID_INTEL, 0x1531, NT, "Intel", "I210 Gigabit Network Connection Unprogrammed"}, + {PCI_VENDOR_ID_INTEL, 0x1532, NT, "Intel", "I211 Gigabit Network Connection Unprogrammed"}, + {PCI_VENDOR_ID_INTEL, 0x1533, OK, "Intel", "I210 Gigabit Network Connection"}, + {PCI_VENDOR_ID_INTEL, 0x1536, NT, "Intel", "I210 Gigabit Network Connection SERDES Fiber"}, + {PCI_VENDOR_ID_INTEL, 0x1537, NT, "Intel", "I210 Gigabit Network Connection SERDES Backplane"}, + {PCI_VENDOR_ID_INTEL, 0x1538, NT, "Intel", "I210 Gigabit Network Connection SGMII"}, + {PCI_VENDOR_ID_INTEL, 0x1539, NT, "Intel", "I211 Gigabit Network Connection"}, {0}, }; -static int nicintel_ee_probe(struct flashctx *flash) +static inline bool is_i210(uint16_t device_id) +{ + return (device_id & 0xff00) == 0x1500; +} + +static int nicintel_ee_probe_i210(struct flashctx *flash) +{ + /* Emulated eeprom has a fixed size of 4 KB */ + flash->chip->total_size = 4; + flash->chip->page_size = flash->chip->total_size * 1024; + flash->chip->tested = TEST_OK_PREW; + flash->chip->gran = write_gran_1byte_implicit_erase; + flash->chip->block_erasers->eraseblocks[0].size = flash->chip->page_size; + flash->chip->block_erasers->eraseblocks[0].count = 1; + + return 1; +} + +static int nicintel_ee_probe_82580(struct flashctx *flash) { if (nicintel_pci->device_id == UNPROG_DEVICE) flash->chip->total_size = 16; /* Fall back to minimum supported size. */ @@ -103,6 +144,15 @@ static int nicintel_ee_probe(struct flashctx *flash) return 1; } +static int nicintel_ee_probe(struct flashctx *flash) +{ + if (is_i210(nicintel_pci->device_id)) + return nicintel_ee_probe_i210(flash); + + return nicintel_ee_probe_82580(flash); +} + +#define MAX_ATEMPTS 10000000 static int nicintel_ee_read_word(unsigned int addr, uint16_t *data) { uint32_t tmp = BIT(EERD_START) | (addr << EERD_ADDR); @@ -110,7 +160,7 @@ static int nicintel_ee_read_word(unsigned int addr, uint16_t *data) /* Poll done flag. 10.000.000 cycles seem to be enough. */ uint32_t i; - for (i = 0; i < 10000000; i++) { + for (i = 0; i < MAX_ATEMPTS; i++) { tmp = pci_mmio_readl(nicintel_eebar + EERD); if (tmp & BIT(EERD_DONE)) { *data = (tmp >> EERD_DATA) & 0xffff; @@ -151,6 +201,83 @@ static int nicintel_ee_read(struct flashctx *flash, uint8_t *buf, unsigned int a return 0; } +static int nicintel_ee_write_word_i210(unsigned int addr, uint16_t data) +{ + uint32_t eewr; + + eewr = addr << EEWR_ADDR; + eewr |= data << EEWR_DATA; + eewr |= BIT(EEWR_CMDV); + pci_mmio_writel(eewr, nicintel_eebar + EEWR); + + programmer_delay(5); + for (int i = 0; i < MAX_ATEMPTS; i++) + if (pci_mmio_readl(nicintel_eebar + EEWR) & BIT(EEWR_DONE)) + return 0; + return -1; +} + +static int nicintel_ee_write_i210(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) +{ + done_i20_write = true; + + if (addr & 1) { + uint16_t data; + + if (nicintel_ee_read_word(addr / 2, &data)) { + msg_perr("Timeout reading heading byte\n"); + return -1; + } + + data &= 0xff; + data |= (buf ? (buf[0]) : 0xff) << 8; + + if (nicintel_ee_write_word_i210(addr / 2, data)) { + msg_perr("Timeout writing heading word\n"); + return -1; + } + + if (buf) + buf ++; + addr ++; + len --; + } + + while (len > 0) { + uint16_t data; + + if (len == 1) { + if (nicintel_ee_read_word(addr / 2, &data)) { + msg_perr("Timeout reading tail byte\n"); + return -1; + } + + data &= 0xff00; + data |= buf ? (buf[0]) : 0xff; + } else { + if (buf) + data = buf[0] | (buf[1] << 8); + else + data = 0xffff; + } + + if (nicintel_ee_write_word_i210(addr / 2, data)) { + msg_perr("Timeout writing Shadow RAM\n"); + return -1; + } + + if (buf) + buf += 2; + if (len > 2) + len -= 2; + else + len = 0; + addr += 2; + } + + return 0; +} + static int nicintel_ee_bitset(int reg, int bit, bool val) { uint32_t tmp; @@ -224,7 +351,7 @@ static int nicintel_ee_req(void) return 0; } -static int nicintel_ee_write(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) +static int nicintel_ee_write_82580(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) { if (nicintel_ee_req()) return -1; @@ -262,6 +389,13 @@ out: nicintel_ee_bitset(EEC, EE_REQ, 0); /* Give up direct access. */ return ret; } +static int nicintel_ee_write(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) +{ + if (is_i210(nicintel_pci->device_id)) + return nicintel_ee_write_i210(flash, buf, addr, len); + + return nicintel_ee_write_82580(flash, buf, addr, len); +} static int nicintel_ee_erase(struct flashctx *flash, unsigned int addr, unsigned int len) { @@ -275,6 +409,25 @@ static const struct opaque_master opaque_master_nicintel_ee = { .erase = nicintel_ee_erase, }; +static int nicintel_ee_shutdown_i210(void *arg) +{ + if (!done_i20_write) + return 0; + + uint32_t flup = pci_mmio_readl(nicintel_eebar + EEC); + + flup |= BIT(EE_FLUPD); + pci_mmio_writel(flup, nicintel_eebar + EEC); + + for (int i = 0; i < MAX_ATEMPTS; i++) + if (pci_mmio_readl(nicintel_eebar + EEC) & BIT(EE_FLUDONE)) + return 0; + + msg_perr("Flash update failed\n"); + + return -1; +} + static int nicintel_ee_shutdown(void *eecp) { uint32_t old_eec = *(uint32_t *)eecp; @@ -306,9 +459,14 @@ int nicintel_ee_init(void) if (!io_base_addr) return 1; - nicintel_eebar = rphysmap("Intel Gigabit NIC w/ SPI EEPROM", io_base_addr, MEMMAP_SIZE); + nicintel_eebar = rphysmap("Intel Gigabit NIC w/ SPI EEPROM", + io_base_addr + (is_i210(dev->device_id) ? 0x12000 : 0), MEMMAP_SIZE); + if (!nicintel_eebar) + return 1; + nicintel_pci = dev; - if (dev->device_id != UNPROG_DEVICE) { + if ((dev->device_id != UNPROG_DEVICE) && ! is_i210(dev->device_id)) + { uint32_t eec = pci_mmio_readl(nicintel_eebar + EEC); /* C.f. 3.3.1.5 for the detection mechanism (maybe? contradicting the EE_PRES definition), @@ -327,5 +485,9 @@ int nicintel_ee_init(void) return 1; } + if (is_i210(dev->device_id)) + if (register_shutdown(nicintel_ee_shutdown_i210, NULL)) + return 1; + return register_opaque_master(&opaque_master_nicintel_ee); } -- 2.11.0

3 years, 10 months

[PATCH] nicintel_eeprom: Support for I210 emulated EEprom

by Ricardo Ribalda Delgado

On the I210 family there is no MAC EEprom, instead there is a big flash (typically around 16Mb) with contents of the old MAC plus other stuff. There is an interface to program the whole flash, but once it is programmed it enters a "Secure Mode" that disables the interface. Luckily, the section with the MAC can still be updated via the EEprom interface. This patch adds support for this interface. root@qt5022-fglrx:~# ./flashrom -p nicintel_eeprom:pci=01:0.0 -w kk.raw -V flashrom v0.9.9-unknown on Linux 4.10.0-qtec-standard (x86_64) flashrom is free software, get the source code at https://flashrom.org flashrom was built with libpci 3.4.1, GCC 5.3.0, little endian Command line (5 args): ./flashrom -p nicintel_eeprom:pci=01:0.0 -w kk.raw -V Calibrating delay loop... OS timer resolution is 1 usecs, 1856M loops per second, 10 myus = 10 us, 100 myus = 102 us, 1000 myus = 1017 us, 10000 myus = 10044 us, 4 myus = 4 us, OK. Initializing nicintel_eeprom programmer Found "Intel I210 Gigabit Network Connection" (8086:1533, BDF 01:00.0). Requested BAR is of type MEM, 32bit, not prefetchable Requested BAR is of type MEM, 32bit, not prefetchable The following protocols are supported: Programmer-specific. Probing for Programmer Opaque flash chip, 0 kB: Found Programmer flash chip "Opaque flash chip" (4 kB, Programmer-specific) on nicintel_eeprom. Found Programmer flash chip "Opaque flash chip" (4 kB, Programmer-specific). Reading old flash chip contents... done. Erasing and writing flash chip... Trying erase function 0... 0x000000-0x000fff:W Erase/write done. Verifying flash... VERIFIED. Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda(a)gmail.com> --- nicintel_eeprom.c | 172 ++++++++++++++++++++++++++++++++++++++++++++++++++++-- 1 file changed, 166 insertions(+), 6 deletions(-) diff --git a/nicintel_eeprom.c b/nicintel_eeprom.c index b5d4202b01e0..77754f0f2534 100644 --- a/nicintel_eeprom.c +++ b/nicintel_eeprom.c @@ -25,6 +25,12 @@ * 4.7: Access to shared resources (FIXME: we should probably use this semaphore interface) * 7.4: Register Descriptions */ +/* + * Datasheet: Intel Ethernet Controller I210: Datasheet + * 8.4.3: EEPROM-Mode Read Register + * 8.4.6: EEPROM-Mode Write Register + * Write process inspired on kernel e1000_i210.c + */ #include <stdlib.h> #include <unistd.h> @@ -34,10 +40,11 @@ #include "hwaccess.h" #define PCI_VENDOR_ID_INTEL 0x8086 -#define MEMMAP_SIZE (0x14 + 3) /* Only EEC and EERD are needed. */ +#define MEMMAP_SIZE 0x1c /* Only EEC, EERD and EEWR are needed. */ #define EEC 0x10 /* EEPROM/Flash Control Register */ #define EERD 0x14 /* EEPROM Read Register */ +#define EEWR 0x18 /* EEPROM Write Register */ /* EPROM/Flash Control Register bits */ #define EE_SCK 0 @@ -49,6 +56,8 @@ #define EE_PRES 8 #define EE_SIZE 11 #define EE_SIZE_MASK 0xf +#define EE_FLUPD 23 +#define EE_FLUDONE 26 /* EEPROM Read Register bits */ #define EERD_START 0 @@ -56,11 +65,18 @@ #define EERD_ADDR 2 #define EERD_DATA 16 +/* EEPROM Write Register bits */ +#define EEWR_CMDV 0 +#define EEWR_DONE 1 +#define EEWR_ADDR 2 +#define EEWR_DATA 16 + #define BIT(x) (1<<x) #define EE_PAGE_MASK 0x3f static uint8_t *nicintel_eebar; static struct pci_dev *nicintel_pci; +static bool done_i20_write = false; #define UNPROG_DEVICE 0x1509 @@ -71,10 +87,35 @@ const struct dev_entry nics_intel_ee[] = { {PCI_VENDOR_ID_INTEL, 0x1511, NT , "Intel", "82580 Quad Gigabit Ethernet Controller (Ext. PHY)"}, {PCI_VENDOR_ID_INTEL, 0x1511, NT , "Intel", "82580 Dual Gigabit Ethernet Controller (Copper)"}, {PCI_VENDOR_ID_INTEL, UNPROG_DEVICE, OK, "Intel", "Unprogrammed 82580 Quad/Dual Gigabit Ethernet Controller"}, + {PCI_VENDOR_ID_INTEL, 0x1531, NT, "Intel", "I210 Gigabit Network Connection Unprogrammed"}, + {PCI_VENDOR_ID_INTEL, 0x1532, NT, "Intel", "I211 Gigabit Network Connection Unprogrammed"}, + {PCI_VENDOR_ID_INTEL, 0x1533, OK, "Intel", "I210 Gigabit Network Connection"}, + {PCI_VENDOR_ID_INTEL, 0x1536, NT, "Intel", "I210 Gigabit Network Connection SERDES Fiber"}, + {PCI_VENDOR_ID_INTEL, 0x1537, NT, "Intel", "I210 Gigabit Network Connection SERDES Backplane"}, + {PCI_VENDOR_ID_INTEL, 0x1538, NT, "Intel", "I210 Gigabit Network Connection SGMII"}, + {PCI_VENDOR_ID_INTEL, 0x1539, NT, "Intel", "I211 Gigabit Network Connection"}, {0}, }; -static int nicintel_ee_probe(struct flashctx *flash) +static inline bool is_i210(uint16_t device_id) +{ + return (device_id & 0xff00) == 0x1500; +} + +static int nicintel_ee_probe_i210(struct flashctx *flash) +{ + /* Emulated eeprom has a fixed size of 4 KB */ + flash->chip->total_size = 4; + flash->chip->page_size = flash->chip->total_size * 1024; + flash->chip->tested = TEST_OK_PREW; + flash->chip->gran = write_gran_1byte_implicit_erase; + flash->chip->block_erasers->eraseblocks[0].size = flash->chip->page_size; + flash->chip->block_erasers->eraseblocks[0].count = 1; + + return 1; +} + +static int nicintel_ee_probe_82580(struct flashctx *flash) { if (nicintel_pci->device_id == UNPROG_DEVICE) flash->chip->total_size = 16; /* Fall back to minimum supported size. */ @@ -103,6 +144,15 @@ static int nicintel_ee_probe(struct flashctx *flash) return 1; } +static int nicintel_ee_probe(struct flashctx *flash) +{ + if (is_i210(nicintel_pci->device_id)) + return nicintel_ee_probe_i210(flash); + + retirm nicintel_ee_probe_82580(flash); +} + +#define MAX_ATEMPTS 10000000 static int nicintel_ee_read_word(unsigned int addr, uint16_t *data) { uint32_t tmp = BIT(EERD_START) | (addr << EERD_ADDR); @@ -110,7 +160,7 @@ static int nicintel_ee_read_word(unsigned int addr, uint16_t *data) /* Poll done flag. 10.000.000 cycles seem to be enough. */ uint32_t i; - for (i = 0; i < 10000000; i++) { + for (i = 0; i < MAX_ATEMPTS; i++) { tmp = pci_mmio_readl(nicintel_eebar + EERD); if (tmp & BIT(EERD_DONE)) { *data = (tmp >> EERD_DATA) & 0xffff; @@ -151,6 +201,80 @@ static int nicintel_ee_read(struct flashctx *flash, uint8_t *buf, unsigned int a return 0; } +static int nicintel_ee_write_word_i210(struct flashctx *flash, unsigned int addr, uint16_t data) +{ + uint32_t eewr; + + eewr = (addr/2) << EEWR_ADDR; + eewr |= data << EEWR_DATA; + eewr |= BIT(EEWR_CMDV); + pci_mmio_writel(eewr, nicintel_eebar + EEWR); + + programmer_delay(5); + for (int i = 0; i < MAX_ATEMPTS; i++) + if (pci_mmio_readl(nicintel_eebar + EEWR) & BIT(EEWR_DONE)) + return 0; + return -1; +} + +static int nicintel_ee_write_i210(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) +{ + done_i20_write = true; + + if (addr & 1) { + uint16_t data; + + if (nicintel_ee_read_word(addr-1, &data)) { + msg_perr("Timeout reading heading byte\n"); + return -1; + } + + data &= 0xff; + data |= (buf ? (buf[0]) : 0xff) << 8; + + if (nicintel_ee_write_word_i210(flash, addr-1, data)) { + msg_perr("Timeout writing heading word\n"); + return -1; + } + + if (buf) + buf ++; + addr ++; + len ++; + } + + while (len > 0) { + uint16_t data; + + if (len == 1) { + if (nicintel_ee_read_word(addr-1, &data)) { + msg_perr("Timeout reading tail byte\n"); + return -1; + } + + data &= 0xff00; + data |= buf ? (buf[0]) : 0xff; + } else { + if (buf) + data = buf[0] | (buf[1] << 8); + else + data = 0xffff; + } + + if (nicintel_ee_write_word_i210(flash, addr, data)) { + msg_perr("Timeout writing Shadow RAM\n"); + return -1; + } + + if (buf) + buf += 2;; + len -= 2; + addr += 2; + } + + return 0; +} + static int nicintel_ee_bitset(int reg, int bit, bool val) { uint32_t tmp; @@ -224,7 +348,7 @@ static int nicintel_ee_req(void) return 0; } -static int nicintel_ee_write(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) +static int nicintel_ee_write_82580(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) { if (nicintel_ee_req()) return -1; @@ -262,6 +386,13 @@ out: nicintel_ee_bitset(EEC, EE_REQ, 0); /* Give up direct access. */ return ret; } +static int nicintel_ee_write(struct flashctx *flash, const uint8_t *buf, unsigned int addr, unsigned int len) +{ + if (is_i210(nicintel_pci->device_id)) + return nicintel_ee_write_i210(flash, buf, addr, len); + + return nicintel_ee_write_82580(flash, buf, addr, len); +} static int nicintel_ee_erase(struct flashctx *flash, unsigned int addr, unsigned int len) { @@ -275,6 +406,25 @@ static const struct opaque_master opaque_master_nicintel_ee = { .erase = nicintel_ee_erase, }; +static int nicintel_ee_shutdown_i210(void *arg) +{ + if (!done_i20_write) + return 0; + + uint32_t flup = pci_mmio_readl(nicintel_eebar + EEC); + + flup |= BIT(EE_FLUPD); + pci_mmio_writel(flup, nicintel_eebar + EEC); + + for (int i = 0; i < MAX_ATEMPTS; i++) + if (pci_mmio_readl(nicintel_eebar + EEC) & BIT(EE_FLUDONE)) + return 0; + + msg_perr("Flash update failed\n"); + + return -1; +} + static int nicintel_ee_shutdown(void *eecp) { uint32_t old_eec = *(uint32_t *)eecp; @@ -298,6 +448,7 @@ int nicintel_ee_init(void) if (rget_io_perms()) return 1; + struct pci_dev *dev = pcidev_init(nics_intel_ee, PCI_BASE_ADDRESS_0); if (!dev) return 1; @@ -306,9 +457,14 @@ int nicintel_ee_init(void) if (!io_base_addr) return 1; - nicintel_eebar = rphysmap("Intel Gigabit NIC w/ SPI EEPROM", io_base_addr, MEMMAP_SIZE); + nicintel_eebar = rphysmap("Intel Gigabit NIC w/ SPI EEPROM", + io_base_addr + (is_i210(dev->device_id) ? 0x12000 : 0), MEMMAP_SIZE); + if (!nicintel_eebar) + return 1; + nicintel_pci = dev; - if (dev->device_id != UNPROG_DEVICE) { + if ((dev->device_id != UNPROG_DEVICE) && ! is_i210(dev->device_id)) + { uint32_t eec = pci_mmio_readl(nicintel_eebar + EEC); /* C.f. 3.3.1.5 for the detection mechanism (maybe? contradicting the EE_PRES definition), @@ -327,5 +483,9 @@ int nicintel_ee_init(void) return 1; } + if (is_i210(dev->device_id)) + if (register_shutdown(nicintel_ee_shutdown_i210, NULL)) + return 1; + return register_opaque_master(&opaque_master_nicintel_ee); } -- 2.11.0

3 years, 10 months

Chip AT26DF321-SU is recognized as AT25DF321

by The_Raven Raven

Hello I discovered that the chip "AT26DF321-SU" is recognized as "AT25DF321". This is the output from a "AT26DF321-SU": Probing for Atmel AT25DF321, 4096 kB: probe_spi_rdid_generic: id1 0x1f, id2 0x4700 Found Atmel flash chip "AT25DF321" (4096 kB, SPI) on serprog. Chip status register is 0x1c. Chip status register: Sector Protection Register Lock (SRPL) is not set Chip status register: Bit 6 is not set Chip status register: Erase/Program Error (EPE) is not set Chip status register: WP# pin (WPP) is not asserted Chip status register: Software Protection Status (SWP): all sectors are protected Chip status register: Write Enable Latch (WEL) is not set Chip status register: Write In Progress (WIP/BUSY) is not set It seems that both chips have the same ID's. I have readed and written the "AT26DF321-SU" and it worked. If you need more informations or if i should test something -> Let me know. Greetings Greetings

3 years, 10 months

Re: [flashrom] Problem with flashrom and FT2232 Port B

by Nico Huber

On 22.03.2017 14:08, Zak wrote: > Heya > > Thanks for the reply, I'm using the latest version 1.2 for windows and I > use this programmer with other softwares with that same dll and it works > fine on both ports > > But in case I tried with the old libftdi 0.18 and it's the same error Sorry, I'm just poking in the dark. Another idea: Some other driver or program might have claimed access to port B and is blocking it. flash- rom doesn't know how to handle that, maybe your "other softwares" do. Nico > > Regards > > Le 22 mars 2017 12:14, "Nico Huber" <nico.huber(a)secunet.com> a écrit : > >> Hi Zak, >> >> On 21.03.2017 17:28, Zak wrote: >>> Hello, >>> >>> I got flashrom compiled for windows and it works like a charm (has to >>> tweet the code a little, add data for the chips i want to work with, add >>> 4BA and fast rading for FT2232), it works perfectly fine but only for >>> the port A (flashrom -p ft2232_spi:type=2232H,port=A) >>> >>> Using default programmer "ft2232_spi" with arguments "type=2232H,port=A". >>> Calibrating delay loop... OK. >>> Found Macronix flash chip "flashrom -p >>> ft2232_spi:type=2232H,port=AMX25L25665E" (32768 kB, SPI) >>> on ft2232_spi. >>> Reading flash... (0% to done.) >>> >>> If i use the port B (flashrom -p ft2232_spi:type=2232H,port=B), i get >>> this error >>> >>> Calibrating delay loop... OK. >>> Unable to reset FTDI device (USB device unavailable). >>> Unable to set latency timer (USB device unavailable). >>> Unable to set bitmode to SPI (USB device unavailable). >>> ftdi_write_data: -666, USB device unavailable >>> Error: Programmer initialization failed. >>> >>> I get this error also if i try to compile with make >>> CONFIG_DEFAULT_PROGRAMMER=PROGRAMMER_FT2232_SPI >>> CONFIG_DEFAULT_PROGRAMMER_ARGS="type=2232H,port=B" to force port B >>> >>> I looked around i didn't find any solution to fix this, anyone could >>> give me a hand? >> >> the problem probably lies within your (windows) version of libftdi. I'd >> try another OS / libftdi to rule that out. >> >> Nico >> >> >

3 years, 10 months

[PATCH] nicintel_spi: Support for I210/I211 cards

by Ricardo Ribalda Delgado

Implements I210 "raw" flash access as detailed in: http://www.intel.com/content/www/us/en/embedded/products/networking/i210-et… Unfortunately, most of the time the card is in Secure Mode, which means that the raw access is not available. But his should be pretty useful for bringing up boards. Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda(a)gmail.com> --- nicintel_spi.c | 94 ++++++++++++++++++++++++++++++++++++++++++++-------------- 1 file changed, 71 insertions(+), 23 deletions(-) diff --git a/nicintel_spi.c b/nicintel_spi.c index 9195c79e56c9..89f162f96e00 100644 --- a/nicintel_spi.c +++ b/nicintel_spi.c @@ -70,6 +70,9 @@ #define FL_SO 3 #define FL_REQ 4 #define FL_GNT 5 +#define FL_LOCKED 6 +#define FL_ABORT 7 +#define FL_CLR_ERR 8 /* Currently unused */ // #define FL_BUSY 30 // #define FL_ER 31 @@ -95,6 +98,14 @@ const struct dev_entry nics_intel_spi[] = { {PCI_VENDOR_ID_INTEL, 0x152a, NT, "Intel", "82599 10 Gigabit Dual Port Backplane Controller with FCoE"}, {PCI_VENDOR_ID_INTEL, 0x1557, NT, "Intel", "82599 10 Gigabit SFI Network Controller"}, + {PCI_VENDOR_ID_INTEL, 0x1531, OK, "Intel", "I210 Gigabit Network Connection Unprogrammed"}, + {PCI_VENDOR_ID_INTEL, 0x1532, NT, "Intel", "I211 Gigabit Network Connection Unprogrammed"}, + {PCI_VENDOR_ID_INTEL, 0x1533, NT, "Intel", "I210 Gigabit Network Connection"}, + {PCI_VENDOR_ID_INTEL, 0x1536, NT, "Intel", "I210 Gigabit Network Connection SERDES Fiber"}, + {PCI_VENDOR_ID_INTEL, 0x1537, NT, "Intel", "I210 Gigabit Network Connection SERDES Backplane"}, + {PCI_VENDOR_ID_INTEL, 0x1538, NT, "Intel", "I210 Gigabit Network Connection SGMII"}, + {PCI_VENDOR_ID_INTEL, 0x1539, NT, "Intel", "I211 Gigabit Network Connection"}, + {0}, }; @@ -182,32 +193,10 @@ static int nicintel_spi_shutdown(void *data) return 0; } -int nicintel_spi_init(void) +static int nicintel_spi_82599_enable_flash() { - struct pci_dev *dev = NULL; uint32_t tmp; - if (rget_io_perms()) - return 1; - - dev = pcidev_init(nics_intel_spi, PCI_BASE_ADDRESS_0); - if (!dev) - return 1; - - uint32_t io_base_addr = pcidev_readbar(dev, PCI_BASE_ADDRESS_0); - if (!io_base_addr) - return 1; - - if (dev->device_id < 0x10d8) { - nicintel_spibar = rphysmap("Intel Gigabit NIC w/ SPI flash", io_base_addr, - MEMMAP_SIZE); - } else { - nicintel_spibar = rphysmap("Intel 10 Gigabit NIC w/ SPI flash", io_base_addr + 0x10000, - MEMMAP_SIZE); - } - if (nicintel_spibar == ERROR_PTR) - return 1; - /* Automatic restore of EECD on shutdown is not possible because EECD * does not only contain FLASH_WRITES_DISABLED|FLASH_WRITES_ENABLED, * but other bits with side effects as well. Those other bits must be @@ -228,6 +217,65 @@ int nicintel_spi_init(void) if (register_shutdown(nicintel_spi_shutdown, NULL)) return 1; + return 0; +} + +static int nicintel_spi_i210_enable_flash() +{ + uint32_t tmp; + + tmp = pci_mmio_readl(nicintel_spibar + FLA); + if (tmp & (1 << FL_LOCKED)) { + msg_perr("Flash is in Secure Mode. Abort.\n"); + return 1; + } + + if (!(tmp & (1 << FL_ABORT))) + return 0; + + tmp |= (1 << FL_CLR_ERR); + pci_mmio_writel(tmp, nicintel_spibar + FLA); + tmp = pci_mmio_readl(nicintel_spibar + FLA); + if (!tmp & (1 << FL_ABORT)) { + msg_perr("Unable to clear Flash Access Error. Abort\n"); + return 1; + } + + return 0; +} + +int nicintel_spi_init(void) +{ + struct pci_dev *dev = NULL; + + if (rget_io_perms()) + return 1; + + dev = pcidev_init(nics_intel_spi, PCI_BASE_ADDRESS_0); + if (!dev) + return 1; + + uint32_t io_base_addr = pcidev_readbar(dev, PCI_BASE_ADDRESS_0); + if (!io_base_addr) + return 1; + + if ((dev->device_id & 0xfff0) == 0x1530) { + nicintel_spibar = rphysmap("Intel I210 Gigabit w/ SPI flash", io_base_addr + 0x12000, + MEMMAP_SIZE); + if (!nicintel_spibar || nicintel_spi_i210_enable_flash()) + return 1; + } else if (dev->device_id < 0x10d8) { + nicintel_spibar = rphysmap("Intel Gigabit NIC w/ SPI flash", io_base_addr, + MEMMAP_SIZE); + if (!nicintel_spibar || nicintel_spi_82599_enable_flash()) + return 1; + } else { + nicintel_spibar = rphysmap("Intel 10 Gigabit NIC w/ SPI flash", io_base_addr + 0x10000, + MEMMAP_SIZE); + if (!nicintel_spibar || nicintel_spi_82599_enable_flash()) + return 1; + } + if (register_spi_bitbang_master(&bitbang_spi_master_nicintel)) return 1; -- 2.11.0

3 years, 10 months

[PATCH] sb600spi: Add support for Merlin Falcon Chipset

by Ricardo Ribalda Delgado

This patch has been tested on a board similar to AMD Bettong. 00:14.0 SMBus [0c05]: Advanced Micro Devices, Inc. [AMD] FCH SMBus Controller [1022:790b] (rev 4a) 00:14.3 ISA bridge [0601]: Advanced Micro Devices, Inc. [AMD] FCH LPC Bridge [1022:790e] (rev 11) root@qt5022-fglrx:~# ./flashrom -p internal -w kk.rom flashrom v0.9.9-unknown on Linux 4.10.0-qtec-standard (x86_64) flashrom is free software, get the source code at https://flashrom.org Calibrating delay loop... OK. coreboot table found at 0x9ffd6000. Found chipset "AMD FP4". Enabling flash write... OK. Found Micron/Numonyx/ST flash chip "N25Q128..1E" (16384 kB, SPI) mapped at physical address 0x00000000ff000000. Reading old flash chip contents... done. Erasing and writing flash chip... Erase/write done. Verifying flash... VERIFIED. Signed-off-by: Ricardo Ribalda Delgado <ricardo.ribalda(a)gmail.com> --- chipset_enable.c | 1 + sb600spi.c | 28 +++++++++++++++++++++++----- 2 files changed, 24 insertions(+), 5 deletions(-) diff --git a/chipset_enable.c b/chipset_enable.c index b181b9349b02..82bac112a29c 100644 --- a/chipset_enable.c +++ b/chipset_enable.c @@ -1492,6 +1492,7 @@ const struct penable chipset_enables[] = { {0x1022, 0x7440, OK, "AMD", "AMD-768", enable_flash_amd_768_8111}, {0x1022, 0x7468, OK, "AMD", "AMD-8111", enable_flash_amd_768_8111}, {0x1022, 0x780e, OK, "AMD", "FCH", enable_flash_sb600}, + {0x1022, 0x790e, OK, "AMD", "FP4", enable_flash_sb600}, {0x1039, 0x0406, NT, "SiS", "501/5101/5501", enable_flash_sis501}, {0x1039, 0x0496, NT, "SiS", "85C496+497", enable_flash_sis85c496}, {0x1039, 0x0530, OK, "SiS", "530", enable_flash_sis530}, diff --git a/sb600spi.c b/sb600spi.c index 6bd56798e839..68f8d8ac532b 100644 --- a/sb600spi.c +++ b/sb600spi.c @@ -149,6 +149,23 @@ static void determine_generation(struct pci_dev *dev) "Please report this to flashrom(a)flashrom.org and include this log and\n" "the output of lspci -nnvx, thanks!.\n", rev); } + } else if (dev->device_id == 0x790e) { + struct pci_dev *smbus_dev = pci_dev_find(0x1022, 0x790B); + if (smbus_dev == NULL) { + msg_pdbg("No SMBus device with ID 1022:790B found.\n"); + return; + } + uint8_t rev = pci_read_byte(smbus_dev, PCI_REVISION_ID); + if (rev == 0x4a) { + amd_gen = CHIPSET_YANGTZE; + msg_pdbg("Yangtze detected.\n"); + } else { + msg_pwarn("FCH device found but SMBus revision 0x%02x does not match known values.\n" + "Please report this to flashrom(a)flashrom.org and include this log and\n" + "the output of lspci -nnvx, thanks!.\n", rev); + } + + #endif } else msg_pwarn("%s: Unknown LPC device %" PRIx16 ":%" PRIx16 ".\n" @@ -646,12 +663,13 @@ int sb600_probe_spi(struct pci_dev *dev) /* Look for the SMBus device. */ smbus_dev = pci_dev_find(0x1002, 0x4385); - if (!smbus_dev) { + if (!smbus_dev) smbus_dev = pci_dev_find(0x1022, 0x780b); /* AMD FCH */ - if (!smbus_dev) { - msg_perr("ERROR: SMBus device not found. Not enabling SPI.\n"); - return ERROR_NONFATAL; - } + if (!smbus_dev) + smbus_dev = pci_dev_find(0x1022, 0x790b); /* AMD FP4 */ + if (!smbus_dev) { + msg_perr("ERROR: SMBus device not found. Not enabling SPI.\n"); + return ERROR_NONFATAL; } /* Note about the bit tests below: If a bit is zero, the GPIO is SPI. */ -- 2.11.0

3 years, 10 months

Problem with flashrom and FT2232 Port B

by Zak

Hello, I got flashrom compiled for windows and it works like a charm (has to tweet the code a little, add data for the chips i want to work with, add 4BA and fast rading for FT2232), it works perfectly fine but only for the port A (flashrom -p ft2232_spi:type=2232H,port=A) Using default programmer "ft2232_spi" with arguments "type=2232H,port=A". Calibrating delay loop... OK. Found Macronix flash chip "flashrom -p ft2232_spi:type=2232H,port=AMX25L25665E" (32768 kB, SPI) on ft2232_spi. Reading flash... (0% to done.) If i use the port B (flashrom -p ft2232_spi:type=2232H,port=B), i get this error Calibrating delay loop... OK. Unable to reset FTDI device (USB device unavailable). Unable to set latency timer (USB device unavailable). Unable to set bitmode to SPI (USB device unavailable). ftdi_write_data: -666, USB device unavailable Error: Programmer initialization failed. I get this error also if i try to compile with make CONFIG_DEFAULT_PROGRAMMER=PROGRAMMER_FT2232_SPI CONFIG_DEFAULT_PROGRAMMER_ARGS="type=2232H,port=B" to force port B I looked around i didn't find any solution to fix this, anyone could give me a hand? Thanks a bunch Regards

3 years, 10 months

Re: [flashrom] Terminology to help distinguish different ways of using Flashrom

by Sam Kuper

On 19/03/2017, David Hendricks <david.hendricks(a)gmail.com> wrote: > On Sun, Mar 19, 2017 at 10:42 AM, Sam Kuper <sam.kuper(a)uclmail.net> wrote: >> Is there a term that unambiguously describes method 2(b)? > > Just to clarify, in that situation is the ROM is in a dedicated programmer > such as http://www.dediprog.com/pd/universal-programmer/progmaster-u4 ? I'd > describe it as "standalone." Good to know. Thank you!

3 years, 10 months

Re: [flashrom] Flashrom wiki: what license is the content under?

by Sam Kuper

On 20/03/2017, Sam Kuper <sam.kuper(a)uclmail.net> wrote: > On 20/03/2017, Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> > wrote: >> On 18.03.2017 17:50, Sam Kuper wrote: >>> Ideally, Flashrom would license the content under the GFDL and CC >>> BY-SA 3.0, making the content entirely license-compatible with content >>> from Wikipedia and from the Stack Exchange network of websites. >> >> Any licensing which inhibits moving code comments to the wiki or vice >> versa is a really bad idea. > > I doubt this would be a problem in practice. I think there are > standard exceptions for this sort of thing. But I'll try to remember > to look into it, because if there *aren't* exceptions for it, then > you're right: it could be a problem. Also, even if the wiki content were licensed under CC BY-SA 3.0, I don't see how there would be any more of an impediment to moving text between the code and the wiki than presently exists. Please correct me if you think I am mistaken.

3 years, 10 months

Re: [flashrom] Flashrom wiki: what license is the content under?

by Carl-Daniel Hailfinger

Hi Sam, welcome to the flashrom mailing list. Starting almost identical discussions on two different mailing lists is not productive. It _might_ make sense to revisit this after the discussion on the coreboot mailing list has come to a conclusion. On 18.03.2017 17:50, Sam Kuper wrote: > I have looked at a number of pages on the Flashrom wiki, though not all of them. > > None of the pages I have looked at mentioned the license (if any) > under which their content is available. > > My understanding is that means that much (maybe all) of the > documentation in the Flashrom wiki is proprietary (at least, in the > overwhelming majority of jurisdictions). IANAL, though. > > [...] > Alternatively, it would be great if the Flashrom wiki contributors > could license the documentation in the wiki under one or more free > culture licenses: https://freedomdefined.org/Licenses > > Ideally, Flashrom would license the content under the GFDL and CC > BY-SA 3.0, making the content entirely license-compatible with content > from Wikipedia and from the Stack Exchange network of websites. Why would we want to do that? The GFDL is pretty much the worst licensing choice: it is GPL incompatible, so we would hurt ourselves by using it. CC licensing might make sense, but I don't see the benefit from making the content license-compatible with wikipedia (anything copied from third-party websites to wikipedia is quickly marked as copyvio, and then reverted), and stackexchange questions about flashrom usually receive zero upvotes (maybe because people tend to look elsewhere for flashrom information). Any licensing which inhibits moving code comments to the wiki or vice versa is a really bad idea. > If you agree with the position I have taken above, please do reply to > this thread to say so, especially if you have suggestions about how to > best achieve the (re-)licensing. > > If you disagree with my position, please reply to explain your disagreement. It looks like you're trying to start a vote with special rules where only those who disagree with you have the burden of explaining their position. I'm afraid that's not how it works. Regards, Carl-Daniel

3 years, 10 months

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

flashrom March 2017