flashrom

flashrom@flashrom.org

8 participants
9164 discussions

Re: [flashrom] [PATCH] Add support for SFDP (JESD216).
by Carl-Daniel Hailfinger 08 Feb '12

08 Feb '12

Am 31.01.2012 06:59 schrieb Stefan Tauner: > Similar to modules using the opaque programmer framework (e.g. ICH Hardware > Sequencing) this uses a template struct flashchip element in flashchips.c with > a special probe function that fills the obtained values into that struct. > > This allows yet unknown SPI chips to be supported (read, erase, write) almost > as if it was already added to flashchips.c. > > Documentation used: > http://www.jedec.org/standards-documents/docs/jesd216 (2011-04) > W25Q32BV data sheet Revision F (2011-04-01) > EN25QH16 data sheet Revision F (2011-06-01) > > todo: > - handle programmers which have a problem with the dummy bytes needed AMD SB[678]x0 SPI has a way to specify sending one dummy byte between write and read, IIRC it is called DropOneClkOnRead or somthing like that. Quite a few other SPI masters have the one-dummy-byte functionality as well. This needs to be implemented in a generic way (I have a totally bitrotted patch for it), but it should not hold back this patch. > - move sfdp_add_uniform_eraser to spi25.c for further use like > spi_get_erasefn_from_opcode? > - is setting the generic SPI unlock method safe? > - what page_size should be set? > > Tested-by: David Hendricks <dhendrix(a)google.com> > on W25Q64CV > > Signed-off-by: Stefan Tauner <stefan.tauner(a)student.tuwien.ac.at> > --- > Makefile | 2 +- > chipdrivers.h | 4 + > flash.h | 2 + > flashchips.c | 23 ++++ > flashchips.h | 1 + > flashrom.c | 28 +++++- > sfdp.c | 355 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ > spi.h | 5 + > spi25.c | 31 +++++ > 9 files changed, 449 insertions(+), 2 deletions(-) > create mode 100644 sfdp.c > > diff --git a/Makefile b/Makefile > index 83aa038..b890d8a 100644 > --- a/Makefile > +++ b/Makefile > @@ -252,7 +252,7 @@ endif > CHIP_OBJS = jedec.o stm50flw0x0x.o w39.o w29ee011.o \ > sst28sf040.o m29f400bt.o 82802ab.o pm49fl00x.o \ > sst49lfxxxc.o sst_fwhub.o flashchips.o spi.o spi25.o sharplhf00l04.o \ > - a25.o at25.o opaque.o > + a25.o at25.o opaque.o sfdp.o > > LIB_OBJS = layout.o > > diff --git a/chipdrivers.h b/chipdrivers.h > index a1d0cd9..bd81098 100644 > --- a/chipdrivers.h > +++ b/chipdrivers.h > @@ -41,6 +41,7 @@ int spi_block_erase_d7(struct flashctx *flash, unsigned int addr, unsigned int b > int spi_block_erase_d8(struct flashctx *flash, unsigned int addr, unsigned int blocklen); > int spi_block_erase_60(struct flashctx *flash, unsigned int addr, unsigned int blocklen); > int spi_block_erase_c7(struct flashctx *flash, unsigned int addr, unsigned int blocklen); > +erasefunc_t *spi_get_erasefn_from_opcode(uint8_t opcode); > int spi_chip_write_1(struct flashctx *flash, uint8_t *buf, unsigned int start, unsigned int len); > int spi_chip_write_256(struct flashctx *flash, uint8_t *buf, unsigned int start, unsigned int len); > int spi_chip_read(struct flashctx *flash, uint8_t *buf, unsigned int start, int unsigned len); > @@ -58,6 +59,9 @@ int spi_read_chunked(struct flashctx *flash, uint8_t *buf, unsigned int start, u > int spi_write_chunked(struct flashctx *flash, uint8_t *buf, unsigned int start, unsigned int len, unsigned int chunksize); > int spi_aai_write(struct flashctx *flash, uint8_t *buf, unsigned int start, unsigned int len); > > +/* sfdp.c */ > +int probe_spi_sfdp(struct flashctx *flash); > + > /* opaque.c */ > int probe_opaque(struct flashctx *flash); > int read_opaque(struct flashctx *flash, uint8_t *buf, unsigned int start, unsigned int len); > diff --git a/flash.h b/flash.h > index e51b6d4..6bcae71 100644 > --- a/flash.h > +++ b/flash.h > @@ -174,6 +174,8 @@ struct flashctx { > struct registered_programmer *pgm; > }; > > +typedef int (erasefunc_t)(struct flashctx *flash, unsigned int addr, unsigned int blocklen); I believe that typedef to be pretty unreadable, but I see that avoiding the typedef would probably be even worse. > + > #define TEST_UNTESTED 0 > > #define TEST_OK_PROBE (1 << 0) > diff --git a/flashchips.c b/flashchips.c > index 0c8257b..b22e709 100644 > --- a/flashchips.c > +++ b/flashchips.c > @@ -8872,6 +8872,29 @@ const struct flashchip flashchips[] = { > .read = read_memmapped, > .voltage = {3000, 3600}, /* Also has 12V fast program */ > }, > + > + { > + .vendor = "Unknown", > + .name = "SFDP device", > + .bustype = BUS_SPI, > + .manufacture_id = GENERIC_MANUF_ID, > + .model_id = SFDP_DEVICE_ID, > + /* We present our own "report this" text hence we do not > + * want the default "This flash part has status UNTESTED..." > + * text to be printed. */ > + .tested = TEST_OK_PREW, > + .probe = probe_spi_sfdp, > + .unlock = spi_disable_blockprotect, /* is this safe? */ Should be safe AFAICS, but that's not a hard statement, it's a gut feeling. > + .read = spi_chip_read, > + .page_size = 256, /* ? */ Argh, page_size comes to bite us again. Did I already send my "kill most uses of page_size" patch? > + /* FIXME: some vendor extensions define this */ > + .voltage = {}, > + /* Everything below will be set by the probing function. */ > + .write = NULL, > + .total_size = 0, > + .feature_bits = 0, > + .block_erasers = {}, > + }, > > { > .vendor = "Unknown", > diff --git a/flashchips.h b/flashchips.h > index 03efb86..1f2a8ca 100644 > --- a/flashchips.h > +++ b/flashchips.h > @@ -36,6 +36,7 @@ > > #define GENERIC_MANUF_ID 0xffff /* Check if there is a vendor ID */ > #define GENERIC_DEVICE_ID 0xffff /* Only match the vendor ID */ > +#define SFDP_DEVICE_ID 0xfffe Side note: Should we move PROGMANUF_ID and its companion from the bottom of the file to this location to have generic match IDs in one place? > > #define ALLIANCE_ID 0x52 /* Alliance Semiconductor */ > #define ALLIANCE_AS29F002B 0x34 > diff --git a/flashrom.c b/flashrom.c > index ee68344..84fb3fc 100644 > --- a/flashrom.c > +++ b/flashrom.c > @@ -986,7 +986,33 @@ int probe_flash(struct registered_programmer *pgm, int startchip, > * probe_flash() is the first one and thus no chip has been > * found before. The comment above is not really valid anymore. Can you include the following snippet in your patch? @@ -980,11 +980,10 @@ /* If this is the first chip found, accept it. * If this is not the first chip found, accept it only if it is - * a non-generic match. - * We could either make chipcount global or provide it as - * parameter, or we assume that startchip==0 means this call to - * probe_flash() is the first one and thus no chip has been - * found before. + * a non-generic match. SFDP and CFI are generic matches. + * startchip==0 means this call to probe_flash() is the first + * one for this programmer interface and thus no other chip has + * been found on this interface. */ if (startchip == 0 || fill_flash->model_id != GENERIC_DEVICE_ID) break; > */ > - if (startchip == 0 || fill_flash->model_id != GENERIC_DEVICE_ID) > + if (startchip == 0 && fill_flash->model_id == SFDP_DEVICE_ID) { > + msg_cinfo("===\n" > + "SFDP has autodetected a flash chip which is " > + "not natively supported by flashrom yet.\n"); > + if (count_usable_erasers(fill_flash) == 0) > + msg_cinfo("The standard operations read and " > + "verify should work, but to support " > + "erase, write and all other " > + "possible features"); > + else > + msg_cinfo("All standard operations (read, " > + "verify, erase and write) should " > + "work, but to support all possible " > + "features"); > + > + msg_cinfo(" we need to add them manually.\nYou " > + "can help us by mailing us the output of " > + "the following command to flashrom@flashrom." > + "org: \n'flashrom -VV [plus the " > + "-p/--programmer parameter (if needed)]" > + "'\nThanks for your help!\n" > + "===\n"); > + } > + > + if (startchip == 0 || > + ((fill_flash->model_id != GENERIC_DEVICE_ID) && > + (fill_flash->model_id != SFDP_DEVICE_ID))) > break; > > notfound: > diff --git a/sfdp.c b/sfdp.c > new file mode 100644 > index 0000000..7c00ff5 > --- /dev/null > +++ b/sfdp.c > @@ -0,0 +1,355 @@ > +/* > + * This file is part of the flashrom project. > + * > + * Copyright (C) 2011-2012 Stefan Tauner > + * > + * This program is free software; you can redistribute it and/or modify > + * it under the terms of the GNU General Public License as published by > + * the Free Software Foundation; version 2 of the License. > + * > + * This program is distributed in the hope that it will be useful, > + * but WITHOUT ANY WARRANTY; without even the implied warranty of > + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the > + * GNU General Public License for more details. > + * > + * You should have received a copy of the GNU General Public License > + * along with this program; if not, write to the Free Software > + * Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA > + */ > + > +#include <stdint.h> > +#include <stdlib.h> > +#include "flash.h" > +#include "spi.h" > +#include "chipdrivers.h" > + > +static int spi_sfdp_wrapper(struct flashctx *flash, uint32_t address, uint8_t *buf, int len) > +{ > + int i, ret; > + const unsigned char cmd[JEDEC_SFDP_OUTSIZE] = { > + JEDEC_SFDP, > + (address >> 16) & 0xff, > + (address >> 8) & 0xff, > + (address >> 0) & 0xff, > + 0 > + }; > + msg_cspew("spi_sfdp_wrapper: addr=0x%x, len=%d, data:\n", address, len); > + ret = spi_send_command(flash, sizeof(cmd), len, cmd, buf); > + for (i = 0; i < len; i++) > + msg_cspew(" 0x%02x", buf[i]); > + msg_cspew("\n"); > + return ret; > +} > + > +/* FIXME: eventually something similar like this but more generic should be > + * available to split up spi commands. use that then instead */ > +static int spi_sfdp(struct flashctx *flash, uint32_t address, uint8_t *buf, int len) > +{ > + /* FIXME: this is wrong. */ > + int maxstep = 8; Yes, maxstep should be a programmer-specific setting. However, with our current infrastructure there is no way to fix this easily. > + int ret = 0; > + while (len > 0) { > + int step = min(len, maxstep); > + ret = spi_sfdp_wrapper(flash, address, buf, step); > + if (ret) Actually, there is probably a way to determine optimal size for those SFDP requests: Check if ret indicates a "command too long" error and reduce maxstep by one in that case, then retry. Such code is not exactly pretty and I'm not sure if all SPI masters have such a differentiated error code reporting. > + return ret; > + address += step; > + buf += step; > + len -= step; > + } > + return ret; > +} > + > +struct sfdp_tbl_hdr { > + uint8_t id; > + uint8_t v_minor; > + uint8_t v_major; > + uint8_t len; > + uint32_t ptp; /* 24b pointer */ > +}; > + > +static int sfdp_add_uniform_eraser(struct flashctx *f, uint8_t opcode, uint32_t bsize) struct flashctx *flash Just "f" is too short, and if you try to search for it, you'll get thousands of matches you don't want. bsize or blocksize? I prefer the latter for consistency and readability reasons. > +{ > + uint32_t total_size = f->total_size * 1024; > + int i; > + erasefunc_t *erasefn = spi_get_erasefn_from_opcode(opcode); > + if (erasefn == NULL) > + return 1; > + for (i = 0; i < NUM_ERASEFUNCTIONS; i++) { > + struct block_eraser *eraser = &f->block_erasers[i]; > + if (eraser->eraseblocks[0].size != 0 || !eraser->block_erase) > + continue; > + eraser->block_erase = erasefn; > + eraser->eraseblocks[0].size = bsize; > + eraser->eraseblocks[0].count = total_size/bsize; > + msg_cdbg2(" Block eraser %d: %d x %d B with opcode " > + "0x%02x\n", i, total_size/bsize, bsize, > + opcode); > + return 0; > + } > + msg_cinfo("%s: Not enough space to store another eraser (i=%d)." msg_cerr? > + " Please report this at flashrom(a)flashrom.org\n", > + __func__, i); > + return 1; > +} > + Rest of review follows later. > +static int sfdp_fill_flash(struct flashctx *f, uint8_t *buf, uint16_t len) > +{ > + uint32_t tmp32; > + uint8_t tmp8; > + uint32_t total_size; /* in bytes */ > + uint32_t bsize; > + uint8_t opcode_4k = 0xFF; > + int dw, j; > + > + msg_cdbg("Parsing JEDEC SFDP parameter table... "); > + if (len != 9 * 4 && len != 4 * 4) { > + msg_cerr("%s: len out of spec\n", __func__); > + return 1; > + } > + msg_cdbg2("\n"); > + > + /* 1. double word */ > + dw = 0; > + tmp32 = buf[(4 * dw) + 0]; > + tmp32 |= ((unsigned int)buf[(4 * dw) + 1]) << 8; > + tmp32 |= ((unsigned int)buf[(4 * dw) + 2]) << 16; > + tmp32 |= ((unsigned int)buf[(4 * dw) + 3]) << 24; > + > + tmp8 = (tmp32 >> 17) & 0x3; > + switch (tmp8) { > + case 0x0: > + msg_cdbg2(" 3-Byte only addressing.\n"); > + break; > + case 0x1: > + msg_cdbg2(" 3-Byte (and optionally 4-Byte) addressing.\n"); > + break; > + case 0x2: > + msg_cdbg(" 4-Byte only addressing not supported.\n"); > + return 1; > + default: > + msg_cdbg(" Required addressing mode (0x%x) not supported.\n", > + tmp8); > + return 1; > + } > + > + msg_cdbg2(" Writes to the status register have "); > + if (tmp32 & (1 << 3)) { > + msg_cdbg2("to be enabled with "); > + if (tmp32 & (1 << 4)) { > + f->feature_bits = FEATURE_WRSR_WREN; > + msg_cdbg2("WREN (0x06).\n"); > + } else { > + f->feature_bits = FEATURE_WRSR_EWSR; > + msg_cdbg2("EWSR (0x50).\n"); > + } > + } else > + msg_cdbg2("not to be especially enabled.\n"); > + > + msg_cdbg2(" Write granularity is "); > + if (tmp32 & (1 << 2)) { > + msg_cdbg2("at least 64 B.\n"); > + f->write = spi_chip_write_256; > + } else { > + msg_cdbg2("1 B only.\n"); > + f->write = spi_chip_write_1; > + } > + > + if ((tmp32 & 0x3) == 0x1) { > + opcode_4k = (tmp32 >> 8) & 0xFF; /* will be dealt with later */ > + } > + > + /* 2. double word */ > + dw = 1; > + tmp32 = buf[(4 * dw) + 0]; > + tmp32 |= ((unsigned int)buf[(4 * dw) + 1]) << 8; > + tmp32 |= ((unsigned int)buf[(4 * dw) + 2]) << 16; > + tmp32 |= ((unsigned int)buf[(4 * dw) + 3]) << 24; > + > + if (tmp32 & (1 << 31)) { > + msg_cerr("Flash chip size >= 4 Gb/512 MB not supported.\n"); > + return 1; > + } > + total_size = ((tmp32 & 0x7FFFFFFF) + 1) / 8; > + f->total_size = total_size / 1024; > + msg_cdbg2(" Flash chip size is %d kB.\n", f->total_size); > + > + /* FIXME: double words 3-7 contain unused fast read information */ > + > + if (len == 4 * 4) { > + msg_cdbg("It seems like this chip supports the preliminary " > + "Intel version of SFDP, skipping processing of double " > + "words 3-9.\n"); > + goto proc_4k; > + } > + > + dw = 8; > + for (j = 0; j < 4; j++) { > + /* 8 double words from the start + 2 words for every eraser */ > + tmp32 = buf[(4 * dw) + (2 * j)]; > + if (tmp32 == 0) { > + msg_cdbg2(" Block eraser %d is unused.\n", j); > + continue; > + } > + if (tmp32 >= 31) { > + msg_cdbg2(" Block size of eraser %d (2^%d) is too big." > + "\n", j, tmp32); > + continue; > + } > + bsize = 1 << (tmp32); /* bsize = 2 ^ field */ > + > + tmp8 = buf[(4 * dw) + (2 * j) + 1]; > + if(sfdp_add_uniform_eraser(f, tmp8, bsize)) > + continue; > + /* If there is a valid 4k value in the last double words, > + * we want to override the value from double word 1, hence force > + * skipping its processing: */ > + if (bsize == 4 * 1024) > + opcode_4k = 0xFF; > + } > + > +proc_4k: > + if (opcode_4k != 0xFF) { > + sfdp_add_uniform_eraser(f, opcode_4k, 4 * 1024); > + } > + msg_cdbg("done.\n"); > + return 0; > +} > + > +static int sfdp_fetch_pt(struct flashctx *flash, uint32_t addr, uint8_t *buf, uint16_t len) > +{ > + uint16_t i; > + if (spi_sfdp(flash, addr, buf, len)) { > + msg_cerr("Receiving SFDP parameter table failed.\n"); > + return 1; > + } > + msg_cspew(" Parameter table contents:\n"); > + for(i = 0; i < len; i++) { > + if ((i % 8) == 0) { > + msg_cspew(" 0x%03x: ", i); > + } > + msg_cspew(" 0x%02x", buf[i]); > + if ((i % 8) == 7) { > + msg_cspew("\n"); > + continue; > + } > + if ((i % 8) == 3) { > + msg_cspew(" "); > + continue; > + } > + } > + msg_cspew("\n"); > + return 0; > +} > + > +int probe_spi_sfdp(struct flashctx *flash) > +{ > + int ret = 0; > + uint8_t buf[8]; > + uint32_t tmp32; > + uint8_t nph; > + /* need to limit the table loop by comparing i to uint8_t nph hence: */ > + uint16_t i; > + struct sfdp_tbl_hdr *hdrs; > + uint8_t *hbuf; > + uint8_t *tbuf; > + > + if (spi_sfdp(flash, 0x00, buf, 4)) { > + msg_cerr("Receiving SFDP signature failed.\n"); > + return 0; > + } > + tmp32 = buf[0]; > + tmp32 |= ((unsigned int)buf[1]) << 8; > + tmp32 |= ((unsigned int)buf[2]) << 16; > + tmp32 |= ((unsigned int)buf[3]) << 24; > + > + msg_cdbg2("SFDP signature = 0x%08x (should be 0x50444653)\n", tmp32); > + if (tmp32 != 0x50444653) { > + msg_cdbg("No SFDP signature found.\n"); > + return 0; > + } > + if (spi_sfdp(flash, 0x04, buf, 3)) { > + msg_cerr("Receiving SFDP revision and number of parameter " > + "headers (NPH) failed. "); > + return 0; > + } > + msg_cdbg2("SFDP revision = %d.%d\n", buf[1], buf[0]); > + nph = buf[2]; > + msg_cdbg2("SFDP number of parameter headers (NPH) = %d (+ 1 mandatory)" > + "\n", nph); > + > + /* Fetch all parameter headers, even if we don't use them all (yet). */ > + hbuf = malloc(sizeof(struct sfdp_tbl_hdr) * (nph + 1)); > + hdrs = malloc((nph + 1) * 8); > + if (hbuf == NULL || hdrs == NULL ) { > + msg_gerr("Out of memory!\n"); > + exit(1); /* FIXME: shutdown gracefully */ > + } > + if (spi_sfdp(flash, 0x08, hbuf, (nph + 1) * 8)) { > + msg_cerr("Receiving SFDP parameter table headers failed.\n"); > + goto cleanup_hdrs; > + } > + > + i = 0; > + do { > + uint16_t len; > + hdrs[i].id = hbuf[(8 * i) + 0]; > + hdrs[i].v_minor = hbuf[(8 * i) + 1]; > + hdrs[i].v_major = hbuf[(8 * i) + 2]; > + hdrs[i].len = hbuf[(8 * i) + 3]; > + hdrs[i].ptp = hbuf[(8 * i) + 4]; > + hdrs[i].ptp |= ((unsigned int)hbuf[(8 * i) + 5]) << 8; > + hdrs[i].ptp |= ((unsigned int)hbuf[(8 * i) + 6]) << 16; > + msg_cdbg2("SFDP parameter table header %d/%d:\n", i, nph); > + msg_cdbg2(" ID 0x%02x, version %d.%d\n", hdrs[i].id, > + hdrs[i].v_major, hdrs[i].v_minor); > + len = hdrs[i].len * 4; > + tmp32 = hdrs[i].ptp; > + msg_cdbg2(" Length %d B, Parameter Table Pointer 0x%06x\n", > + len, tmp32); > + > + if (len + tmp32 > UINT16_MAX) { > + msg_cerr("SFDP Parameter Table %d supposedly " > + "overflows addressable SFDP area. This most\n" > + "probably indicates a corrupt SFDP parameter " > + "table header. Aborting SFDP probe!\n", i); > + ret = 0; > + goto cleanup_hdrs; > + } > + > + tbuf = malloc(len); > + if (tbuf == NULL) { > + msg_gerr("Out of memory!\n"); > + exit(1); /* FIXME: shutdown gracefully */ > + } > + if (sfdp_fetch_pt(flash, tmp32, tbuf, len)){ > + msg_cerr("Fetching SFDP parameter table %d failed.\n", > + i); > + free(tbuf); > + break; > + } > + if (i == 0) { /* Mandatory JEDEC SFDP parameter table */ > + if (hdrs[i].id != 0) > + msg_cdbg("ID of the mandatory JEDEC SFDP " > + "parameter table is not 0 as demanded " > + "by JESD216 (warning only).\n"); > + > + if (len != 9 * 4 && len != 4 * 4) { > + msg_cdbg("Length of the mandatory JEDEC SFDP " > + "parameter table is %d B instead of " > + "36 B (i.e. 9 double words) as " > + "demanded by JESD216, skipping " > + "parsing.\n", len); > + } else if (sfdp_fill_flash(flash, tbuf, len) == 0) > + ret = 1; > + } > + > + free(tbuf); > + i++; > + } while(i <= nph); > + > +cleanup_hdrs: > + free(hdrs); > + free(hbuf); > + return ret; > +} Regards, Carl-Daniel -- http://www.hailfinger.org/

2 3

Re: [flashrom] FAILED : mcp61sm -am
by Joshua Roys 08 Feb '12

08 Feb '12

On 02/08/2012 02:47 PM, Vince S wrote: > can i restore my bios from my linux session or do i have to reboot (i > have another bios with i can reboot) to flash it from a freedos live usb? > > Hello, Please do NOT reboot. Can you provide the log of when flashrom failed? Also, if you made a backup, you can see if your BIOS is still intact by running `flashrom -v /path/to/backup.foo` Thanks, Josh

1 0

FAILED : mcp61sm -am
by Vince S 08 Feb '12

08 Feb '12

can i restore my bios from my linux session or do i have to reboot (i have another bios with i can reboot) to flash it from a freedos live usb?

1 0

[PATCH] dummy programmer: blacklist SPI commands
by Carl-Daniel Hailfinger 07 Feb '12

07 Feb '12

Add optional SPI command blacklisting to the flash chip emulator in the dummy programmer. Usage: flashrom -p dummy:spi_blacklist=hexstring If hexstring is 0302, flashrom will blacklist command 0x03 (READ) and command 0x02 (WRITE). The hexstring can have up to 521 bytes (256 commands) length, and must not start with 0x. TODO: Should 0x at the beginning of hexstring be tolerated? Very useful for testing corner cases if you don't own a locked down Intel chipset and want to simulate such a thing. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Index: flashrom-emulate_spi_flashchip_command_blacklist/dummyflasher.c =================================================================== --- flashrom-emulate_spi_flashchip_command_blacklist/dummyflasher.c (Revision 1224) +++ flashrom-emulate_spi_flashchip_command_blacklist/dummyflasher.c (Arbeitskopie) @@ -19,6 +19,7 @@ #include <string.h> #include <stdlib.h> +#include <stdio.h> #include <ctype.h> #include "flash.h" #include "chipdrivers.h" @@ -56,6 +57,8 @@ static int emu_jedec_be_d8_size = 0; static int emu_jedec_ce_60_size = 0; static int emu_jedec_ce_c7_size = 0; +unsigned char spi_blacklist[256]; +int spi_blacklist_size = 0; #endif #endif @@ -71,6 +74,7 @@ { char *bustext = NULL; char *tmp = NULL; + int i; #if EMULATE_CHIP struct stat image_stat; #endif @@ -116,6 +120,31 @@ } } + tmp = extract_programmer_param("spi_blacklist"); + if (tmp) { + i = strlen(tmp); + if (i % 2) { + msg_perr("Invalid SPI command blacklist\n"); + free(tmp); + return 1; + } + tolower_string(tmp); + if (strspn(tmp, "0123456789abcdef") != i) { + msg_perr("Invalid SPI command blacklist\n"); + free(tmp); + return 1; + } + spi_blacklist_size = i / 2; + for (i = 0; i < spi_blacklist_size; i++) { + sscanf(tmp, "%2hhx", &spi_blacklist[i]); + } + msg_pdbg("SPI blacklist is "); + for (i = 0; i < spi_blacklist_size; i++) + msg_pdbg("%02x ", spi_blacklist[i]); + msg_pdbg(", size %i\n", spi_blacklist_size); + } + free(tmp); + #if EMULATE_CHIP tmp = extract_programmer_param("emulate"); if (!tmp) { @@ -285,7 +314,7 @@ static int emulate_spi_chip_response(unsigned int writecnt, unsigned int readcnt, const unsigned char *writearr, unsigned char *readarr) { - int offs; + int offs, i; static int aai_offs; static int aai_active = 0; @@ -293,7 +322,16 @@ msg_perr("No command sent to the chip!\n"); return 1; } - /* TODO: Implement command blacklists here. */ + for (i = 0; i < spi_blacklist_size; i++) { + if (writearr[0] == spi_blacklist[i]) { + msg_perr("Refusing blacklisted SPI command 0x%02x\n", + spi_blacklist[i]); + /* FIXME: Do we want a separate return code for + * command unavailable? + */ + return 1; + } + } switch (writearr[0]) { case JEDEC_RES: if (emu_chip != EMULATE_ST_M25P10_RES) -- http://www.hailfinger.org/

3 7

[PATCH] Add RDC R6030 chipset
by Rudolf Marek 07 Feb '12

07 Feb '12

Hello, Following patch adds the support for ROMCS_EN for writes aka chipset_enable on RDC R6030 found on bifferboard. The patch for flashchip will follow in near future. Signed-off-by: Rudolf Marek <r.marek(a)assembler.cz> Thanks Rudolf

2 6

[commit] r1489 - trunk
by repository service 07 Feb '12

07 Feb '12

Author: stefanct Date: Tue Feb 7 22:29:48 2012 New Revision: 1489 URL: http://flashrom.org/trac/flashrom/changeset/1489 Log: Add support for RDC R6030 chipset Tested on Bifferboard. Signed-off-by: Rudolf Marek <r.marek(a)assembler.cz> Acked-by: Stefan Tauner <stefan.tauner(a)student.tuwien.ac.at> Modified: trunk/chipset_enable.c trunk/print.c Modified: trunk/chipset_enable.c ============================================================================== --- trunk/chipset_enable.c Fri Feb 3 23:32:09 2012 (r1488) +++ trunk/chipset_enable.c Tue Feb 7 22:29:48 2012 (r1489) @@ -54,6 +54,32 @@ return 0; } +static int enable_flash_rdc_r8610(struct pci_dev *dev, const char *name) +{ + uint8_t tmp; + + /* enable ROMCS for writes */ + tmp = pci_read_byte(dev, 0x43); + tmp |= 0x80; + pci_write_byte(dev, 0x43, tmp); + + /* read the bootstrapping register */ + tmp = pci_read_byte(dev, 0x40) & 0x3; + switch (tmp) { + case 3: + internal_buses_supported = BUS_FWH; + break; + case 2: + internal_buses_supported = BUS_LPC; + break; + default: + internal_buses_supported = BUS_PARALLEL; + break; + } + + return 0; +} + static int enable_flash_sis85c496(struct pci_dev *dev, const char *name) { uint8_t tmp; @@ -1235,6 +1261,7 @@ {0x1106, 0x8409, OK, "VIA", "VX855/VX875", enable_flash_vt823x}, {0x1166, 0x0200, OK, "Broadcom", "OSB4", enable_flash_osb4}, {0x1166, 0x0205, OK, "Broadcom", "HT-1000", enable_flash_ht1000}, + {0x17f3, 0x6030, OK, "RDC", "R8610/R3210", enable_flash_rdc_r8610}, {0x8086, 0x122e, OK, "Intel", "PIIX", enable_flash_piix4}, {0x8086, 0x1234, NT, "Intel", "MPIIX", enable_flash_piix4}, {0x8086, 0x1c44, OK, "Intel", "Z68", enable_flash_pch6}, Modified: trunk/print.c ============================================================================== --- trunk/print.c Fri Feb 3 23:32:09 2012 (r1488) +++ trunk/print.c Tue Feb 7 22:29:48 2012 (r1489) @@ -692,6 +692,7 @@ B("ASUS", "Z8NA-D6C", 1, "http://www.asus.com/Server_Workstation/Server_Motherboards/Z8NAD6C/", NULL), B("ASUS", "Z8PE-D12", 1, "http://www.asus.com/Server_Workstation/Server_Motherboards/Z8PED12/", NULL), B("BCOM", "WinNET100", 1, "http://www.coreboot.org/BCOM_WINNET100", "Used in the IGEL-316 thin client."), + B("Bifferos", "Bifferboard", 1, "http://bifferos.co.uk/", NULL), B("Biostar", "N68S3+", 1, NULL, NULL), B("Biostar", "M6TBA", 0, "ftp://ftp.biostar-usa.com/manuals/M6TBA/", "No public report found. Owned by Uwe Hermann <uwe(a)hermann-uwe.de>. May work now."), B("Biostar", "M7NCD Pro", 1, "http://www.biostar.com.tw/app/en/mb/content.php?S_ID=260", NULL),

1 0

MS-7548 mainboard works
by Richard Albright 07 Feb '12

07 Feb '12

flashed a hp pavilion a6750f successfully -- *Rick Albright* Senior Quantitative Analyst Web: www.insiderscore.com <https://www.insiderscore.com/> Email: ralbright(a)insiderscore.com <mailto:ralbright@insiderscore.com>

2 1

Re: [flashrom] Hello, flashrom problem
by Stefan Tauner 07 Feb '12

07 Feb '12

On Tue, 7 Feb 2012 10:43:51 +0200 Elena Gusanu <elena.gusanu(a)gmail.com> wrote: > I tried but it doesn't work. Hello Elena, Carl-Daniel had the impression that you did just a non-writing run with flashrom (i.e. calling flashrom without any of the -w and -E parameters). It seems that you have done more permanent harm and we need to know more details to be able to help (if at all). What hardware/laptop are we talking about? Which version of flashrom did you use (which distribution? Which command line arguments did you use? Can you remember any error messages it showed when you ran it? -- Kind regards/Mit freundlichen Grüßen, Stefan Tauner

1 0

Re: [flashrom] Shuttle SN78S flash info
by Stefan Tauner 07 Feb '12

07 Feb '12

On Mon, 6 Feb 2012 21:42:56 -0500 Dan Stoner <danstoner(a)gmail.com> wrote: > flashrom v0.9.1-r946 Hi Dan! Please upgrade! Your version of flashrom is 2,5 years old. Nvidia support is still not complete, but it should work way better than with 0.9.1. -- Kind regards/Mit freundlichen Grüßen, Stefan Tauner

2 1

Re: [flashrom] Hello, flashrom problem
by Carl-Daniel Hailfinger 07 Feb '12

07 Feb '12

Hi Elena, Am 06.02.2012 23:54 schrieb Elena Gusanu: > Good morning, > My name is Elena Gusanu and i have a big problem with my laptop. I > flashrom my laptop and I didn't know what was this and i turned off, > unfortunately. Is there a solution to solve my problem and to get back bios? Please remove the power cable and battery from your laptop, wait 1 hour, then plug in battery and power cable and switch on your laptop again. That should work. Regards, Carl-Daniel

2 1

Jump to page:

2025

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

flashrom