diff options
Diffstat (limited to 'drivers/mtd/chips/cfi_cmdset_0001.c')
-rw-r--r-- | drivers/mtd/chips/cfi_cmdset_0001.c | 484 |
1 files changed, 289 insertions, 195 deletions
diff --git a/drivers/mtd/chips/cfi_cmdset_0001.c b/drivers/mtd/chips/cfi_cmdset_0001.c index c3fc9b2f21fb..143f01a4c170 100644 --- a/drivers/mtd/chips/cfi_cmdset_0001.c +++ b/drivers/mtd/chips/cfi_cmdset_0001.c @@ -4,9 +4,9 @@ * * (C) 2000 Red Hat. GPL'd * - * $Id: cfi_cmdset_0001.c,v 1.178 2005/05/19 17:05:43 nico Exp $ + * $Id: cfi_cmdset_0001.c,v 1.185 2005/11/07 11:14:22 gleixner Exp $ + * * - * * 10/10/2000 Nicolas Pitre <nico@cam.org> * - completely revamped method functions so they are aware and * independent of the flash geometry (buswidth, interleave, etc.) @@ -51,6 +51,7 @@ static int cfi_intelext_read (struct mtd_info *, loff_t, size_t, size_t *, u_char *); static int cfi_intelext_write_words(struct mtd_info *, loff_t, size_t, size_t *, const u_char *); static int cfi_intelext_write_buffers(struct mtd_info *, loff_t, size_t, size_t *, const u_char *); +static int cfi_intelext_writev(struct mtd_info *, const struct kvec *, unsigned long, loff_t, size_t *); static int cfi_intelext_erase_varsize(struct mtd_info *, struct erase_info *); static void cfi_intelext_sync (struct mtd_info *); static int cfi_intelext_lock(struct mtd_info *mtd, loff_t ofs, size_t len); @@ -105,6 +106,7 @@ static struct mtd_chip_driver cfi_intelext_chipdrv = { static void cfi_tell_features(struct cfi_pri_intelext *extp) { int i; + printk(" Extended Query version %c.%c\n", extp->MajorVersion, extp->MinorVersion); printk(" Feature/Command Support: %4.4X\n", extp->FeatureSupport); printk(" - Chip Erase: %s\n", extp->FeatureSupport&1?"supported":"unsupported"); printk(" - Suspend Erase: %s\n", extp->FeatureSupport&2?"supported":"unsupported"); @@ -116,36 +118,43 @@ static void cfi_tell_features(struct cfi_pri_intelext *extp) printk(" - Page-mode read: %s\n", extp->FeatureSupport&128?"supported":"unsupported"); printk(" - Synchronous read: %s\n", extp->FeatureSupport&256?"supported":"unsupported"); printk(" - Simultaneous operations: %s\n", extp->FeatureSupport&512?"supported":"unsupported"); - for (i=10; i<32; i++) { - if (extp->FeatureSupport & (1<<i)) + printk(" - Extended Flash Array: %s\n", extp->FeatureSupport&1024?"supported":"unsupported"); + for (i=11; i<32; i++) { + if (extp->FeatureSupport & (1<<i)) printk(" - Unknown Bit %X: supported\n", i); } - + printk(" Supported functions after Suspend: %2.2X\n", extp->SuspendCmdSupport); printk(" - Program after Erase Suspend: %s\n", extp->SuspendCmdSupport&1?"supported":"unsupported"); for (i=1; i<8; i++) { if (extp->SuspendCmdSupport & (1<<i)) printk(" - Unknown Bit %X: supported\n", i); } - + printk(" Block Status Register Mask: %4.4X\n", extp->BlkStatusRegMask); printk(" - Lock Bit Active: %s\n", extp->BlkStatusRegMask&1?"yes":"no"); - printk(" - Valid Bit Active: %s\n", extp->BlkStatusRegMask&2?"yes":"no"); - for (i=2; i<16; i++) { + printk(" - Lock-Down Bit Active: %s\n", extp->BlkStatusRegMask&2?"yes":"no"); + for (i=2; i<3; i++) { if (extp->BlkStatusRegMask & (1<<i)) printk(" - Unknown Bit %X Active: yes\n",i); } - - printk(" Vcc Logic Supply Optimum Program/Erase Voltage: %d.%d V\n", + printk(" - EFA Lock Bit: %s\n", extp->BlkStatusRegMask&16?"yes":"no"); + printk(" - EFA Lock-Down Bit: %s\n", extp->BlkStatusRegMask&32?"yes":"no"); + for (i=6; i<16; i++) { + if (extp->BlkStatusRegMask & (1<<i)) + printk(" - Unknown Bit %X Active: yes\n",i); + } + + printk(" Vcc Logic Supply Optimum Program/Erase Voltage: %d.%d V\n", extp->VccOptimal >> 4, extp->VccOptimal & 0xf); if (extp->VppOptimal) - printk(" Vpp Programming Supply Optimum Program/Erase Voltage: %d.%d V\n", + printk(" Vpp Programming Supply Optimum Program/Erase Voltage: %d.%d V\n", extp->VppOptimal >> 4, extp->VppOptimal & 0xf); } #endif #ifdef CMDSET0001_DISABLE_ERASE_SUSPEND_ON_WRITE -/* Some Intel Strata Flash prior to FPO revision C has bugs in this area */ +/* Some Intel Strata Flash prior to FPO revision C has bugs in this area */ static void fixup_intel_strataflash(struct mtd_info *mtd, void* param) { struct map_info *map = mtd->priv; @@ -176,7 +185,7 @@ static void fixup_st_m28w320ct(struct mtd_info *mtd, void* param) { struct map_info *map = mtd->priv; struct cfi_private *cfi = map->fldrv_priv; - + cfi->cfiq->BufWriteTimeoutTyp = 0; /* Not supported */ cfi->cfiq->BufWriteTimeoutMax = 0; /* Not supported */ } @@ -185,7 +194,7 @@ static void fixup_st_m28w320cb(struct mtd_info *mtd, void* param) { struct map_info *map = mtd->priv; struct cfi_private *cfi = map->fldrv_priv; - + /* Note this is done after the region info is endian swapped */ cfi->cfiq->EraseRegionInfo[1] = (cfi->cfiq->EraseRegionInfo[1] & 0xffff0000) | 0x3e; @@ -207,12 +216,13 @@ static void fixup_use_write_buffers(struct mtd_info *mtd, void *param) if (cfi->cfiq->BufWriteTimeoutTyp) { printk(KERN_INFO "Using buffer write method\n" ); mtd->write = cfi_intelext_write_buffers; + mtd->writev = cfi_intelext_writev; } } static struct cfi_fixup cfi_fixup_table[] = { #ifdef CMDSET0001_DISABLE_ERASE_SUSPEND_ON_WRITE - { CFI_MFR_ANY, CFI_ID_ANY, fixup_intel_strataflash, NULL }, + { CFI_MFR_ANY, CFI_ID_ANY, fixup_intel_strataflash, NULL }, #endif #ifdef CMDSET0001_DISABLE_WRITE_SUSPEND { CFI_MFR_ANY, CFI_ID_ANY, fixup_no_write_suspend, NULL }, @@ -252,12 +262,21 @@ read_pri_intelext(struct map_info *map, __u16 adr) if (!extp) return NULL; + if (extp->MajorVersion != '1' || + (extp->MinorVersion < '0' || extp->MinorVersion > '4')) { + printk(KERN_ERR " Unknown Intel/Sharp Extended Query " + "version %c.%c.\n", extp->MajorVersion, + extp->MinorVersion); + kfree(extp); + return NULL; + } + /* Do some byteswapping if necessary */ extp->FeatureSupport = le32_to_cpu(extp->FeatureSupport); extp->BlkStatusRegMask = le16_to_cpu(extp->BlkStatusRegMask); extp->ProtRegAddr = le16_to_cpu(extp->ProtRegAddr); - if (extp->MajorVersion == '1' && extp->MinorVersion == '3') { + if (extp->MajorVersion == '1' && extp->MinorVersion >= '3') { unsigned int extra_size = 0; int nb_parts, i; @@ -266,7 +285,10 @@ read_pri_intelext(struct map_info *map, __u16 adr) sizeof(struct cfi_intelext_otpinfo); /* Burst Read info */ - extra_size += 6; + extra_size += 2; + if (extp_size < sizeof(*extp) + extra_size) + goto need_more; + extra_size += extp->extra[extra_size-1]; /* Number of hardware-partitions */ extra_size += 1; @@ -274,6 +296,10 @@ read_pri_intelext(struct map_info *map, __u16 adr) goto need_more; nb_parts = extp->extra[extra_size - 1]; + /* skip the sizeof(partregion) field in CFI 1.4 */ + if (extp->MinorVersion >= '4') + extra_size += 2; + for (i = 0; i < nb_parts; i++) { struct cfi_intelext_regioninfo *rinfo; rinfo = (struct cfi_intelext_regioninfo *)&extp->extra[extra_size]; @@ -285,6 +311,9 @@ read_pri_intelext(struct map_info *map, __u16 adr) * sizeof(struct cfi_intelext_blockinfo); } + if (extp->MinorVersion >= '4') + extra_size += sizeof(struct cfi_intelext_programming_regioninfo); + if (extp_size < sizeof(*extp) + extra_size) { need_more: extp_size = sizeof(*extp) + extra_size; @@ -298,7 +327,7 @@ read_pri_intelext(struct map_info *map, __u16 adr) goto again; } } - + return extp; } @@ -339,7 +368,7 @@ struct mtd_info *cfi_cmdset_0001(struct map_info *map, int primary) mtd->reboot_notifier.notifier_call = cfi_intelext_reboot; if (cfi->cfi_mode == CFI_MODE_CFI) { - /* + /* * It's a real CFI chip, not one for which the probe * routine faked a CFI structure. So we read the feature * table from it. @@ -354,14 +383,14 @@ struct mtd_info *cfi_cmdset_0001(struct map_info *map, int primary) } /* Install our own private info structure */ - cfi->cmdset_priv = extp; + cfi->cmdset_priv = extp; cfi_fixup(mtd, cfi_fixup_table); #ifdef DEBUG_CFI_FEATURES /* Tell the user about it in lots of lovely detail */ cfi_tell_features(extp); -#endif +#endif if(extp->SuspendCmdSupport & 1) { printk(KERN_NOTICE "cfi_cmdset_0001: Erase suspend on write enabled\n"); @@ -379,10 +408,10 @@ struct mtd_info *cfi_cmdset_0001(struct map_info *map, int primary) cfi->chips[i].buffer_write_time = 1<<cfi->cfiq->BufWriteTimeoutTyp; cfi->chips[i].erase_time = 1<<cfi->cfiq->BlockEraseTimeoutTyp; cfi->chips[i].ref_point_counter = 0; - } + } map->fldrv = &cfi_intelext_chipdrv; - + return cfi_intelext_setup(mtd); } @@ -399,13 +428,13 @@ static struct mtd_info *cfi_intelext_setup(struct mtd_info *mtd) mtd->size = devsize * cfi->numchips; mtd->numeraseregions = cfi->cfiq->NumEraseRegions * cfi->numchips; - mtd->eraseregions = kmalloc(sizeof(struct mtd_erase_region_info) + mtd->eraseregions = kmalloc(sizeof(struct mtd_erase_region_info) * mtd->numeraseregions, GFP_KERNEL); - if (!mtd->eraseregions) { + if (!mtd->eraseregions) { printk(KERN_ERR "Failed to allocate memory for MTD erase region info\n"); goto setup_err; } - + for (i=0; i<cfi->cfiq->NumEraseRegions; i++) { unsigned long ernum, ersize; ersize = ((cfi->cfiq->EraseRegionInfo[i] >> 8) & ~0xff) * cfi->interleave; @@ -429,7 +458,7 @@ static struct mtd_info *cfi_intelext_setup(struct mtd_info *mtd) } for (i=0; i<mtd->numeraseregions;i++){ - printk(KERN_DEBUG "%d: offset=0x%x,size=0x%x,blocks=%d\n", + printk(KERN_DEBUG "erase region %d: offset=0x%x,size=0x%x,blocks=%d\n", i,mtd->eraseregions[i].offset, mtd->eraseregions[i].erasesize, mtd->eraseregions[i].numblocks); @@ -480,7 +509,7 @@ static int cfi_intelext_partition_fixup(struct mtd_info *mtd, * arrangement at this point. This can be rearranged in the future * if someone feels motivated enough. --nico */ - if (extp && extp->MajorVersion == '1' && extp->MinorVersion == '3' + if (extp && extp->MajorVersion == '1' && extp->MinorVersion >= '3' && extp->FeatureSupport & (1 << 9)) { struct cfi_private *newcfi; struct flchip *chip; @@ -492,12 +521,16 @@ static int cfi_intelext_partition_fixup(struct mtd_info *mtd, sizeof(struct cfi_intelext_otpinfo); /* Burst Read info */ - offs += 6; + offs += extp->extra[offs+1]+2; /* Number of partition regions */ numregions = extp->extra[offs]; offs += 1; + /* skip the sizeof(partregion) field in CFI 1.4 */ + if (extp->MinorVersion >= '4') + offs += 2; + /* Number of hardware partitions */ numparts = 0; for (i = 0; i < numregions; i++) { @@ -509,6 +542,20 @@ static int cfi_intelext_partition_fixup(struct mtd_info *mtd, sizeof(struct cfi_intelext_blockinfo); } + /* Programming Region info */ + if (extp->MinorVersion >= '4') { + struct cfi_intelext_programming_regioninfo *prinfo; + prinfo = (struct cfi_intelext_programming_regioninfo *)&extp->extra[offs]; + MTD_PROGREGION_SIZE(mtd) = cfi->interleave << prinfo->ProgRegShift; + MTD_PROGREGION_CTRLMODE_VALID(mtd) = cfi->interleave * prinfo->ControlValid; + MTD_PROGREGION_CTRLMODE_INVALID(mtd) = cfi->interleave * prinfo->ControlInvalid; + mtd->flags |= MTD_PROGRAM_REGIONS; + printk(KERN_DEBUG "%s: program region size/ctrl_valid/ctrl_inval = %d/%d/%d\n", + map->name, MTD_PROGREGION_SIZE(mtd), + MTD_PROGREGION_CTRLMODE_VALID(mtd), + MTD_PROGREGION_CTRLMODE_INVALID(mtd)); + } + /* * All functions below currently rely on all chips having * the same geometry so we'll just assume that all hardware @@ -653,8 +700,8 @@ static int get_chip(struct map_info *map, struct flchip *chip, unsigned long adr break; if (time_after(jiffies, timeo)) { - printk(KERN_ERR "Waiting for chip to be ready timed out. Status %lx\n", - status.x[0]); + printk(KERN_ERR "%s: Waiting for chip to be ready timed out. Status %lx\n", + map->name, status.x[0]); return -EIO; } spin_unlock(chip->mutex); @@ -663,7 +710,7 @@ static int get_chip(struct map_info *map, struct flchip *chip, unsigned long adr /* Someone else might have been playing with it. */ goto retry; } - + case FL_READY: case FL_CFI_QUERY: case FL_JEDEC_QUERY: @@ -701,8 +748,8 @@ static int get_chip(struct map_info *map, struct flchip *chip, unsigned long adr map_write(map, CMD(0x70), adr); chip->state = FL_ERASING; chip->oldstate = FL_READY; - printk(KERN_ERR "Chip not ready after erase " - "suspended: status = 0x%lx\n", status.x[0]); + printk(KERN_ERR "%s: Chip not ready after erase " + "suspended: status = 0x%lx\n", map->name, status.x[0]); return -EIO; } @@ -782,14 +829,14 @@ static void put_chip(struct map_info *map, struct flchip *chip, unsigned long ad switch(chip->oldstate) { case FL_ERASING: chip->state = chip->oldstate; - /* What if one interleaved chip has finished and the + /* What if one interleaved chip has finished and the other hasn't? The old code would leave the finished - one in READY mode. That's bad, and caused -EROFS + one in READY mode. That's bad, and caused -EROFS errors to be returned from do_erase_oneblock because that's the only bit it checked for at the time. - As the state machine appears to explicitly allow + As the state machine appears to explicitly allow sending the 0x70 (Read Status) command to an erasing - chip and expecting it to be ignored, that's what we + chip and expecting it to be ignored, that's what we do. */ map_write(map, CMD(0xd0), adr); map_write(map, CMD(0x70), adr); @@ -809,7 +856,7 @@ static void put_chip(struct map_info *map, struct flchip *chip, unsigned long ad DISABLE_VPP(map); break; default: - printk(KERN_ERR "put_chip() called with oldstate %d!!\n", chip->oldstate); + printk(KERN_ERR "%s: put_chip() called with oldstate %d!!\n", map->name, chip->oldstate); } wake_up(&chip->wq); } @@ -1025,8 +1072,8 @@ static int do_point_onechip (struct map_info *map, struct flchip *chip, loff_t a adr += chip->start; - /* Ensure cmd read/writes are aligned. */ - cmd_addr = adr & ~(map_bankwidth(map)-1); + /* Ensure cmd read/writes are aligned. */ + cmd_addr = adr & ~(map_bankwidth(map)-1); spin_lock(chip->mutex); @@ -1054,7 +1101,7 @@ static int cfi_intelext_point (struct mtd_info *mtd, loff_t from, size_t len, si if (!map->virt || (from + len > mtd->size)) return -EINVAL; - + *mtdbuf = (void *)map->virt + from; *retlen = 0; @@ -1081,7 +1128,7 @@ static int cfi_intelext_point (struct mtd_info *mtd, loff_t from, size_t len, si *retlen += thislen; len -= thislen; - + ofs = 0; chipnum++; } @@ -1120,7 +1167,7 @@ static void cfi_intelext_unpoint (struct mtd_info *mtd, u_char *addr, loff_t fro if(chip->ref_point_counter == 0) chip->state = FL_READY; } else - printk(KERN_ERR "Warning: unpoint called on non pointed region\n"); /* Should this give an error? */ + printk(KERN_ERR "%s: Warning: unpoint called on non pointed region\n", map->name); /* Should this give an error? */ put_chip(map, chip, chip->start); spin_unlock(chip->mutex); @@ -1139,8 +1186,8 @@ static inline int do_read_onechip(struct map_info *map, struct flchip *chip, lof adr += chip->start; - /* Ensure cmd read/writes are aligned. */ - cmd_addr = adr & ~(map_bankwidth(map)-1); + /* Ensure cmd read/writes are aligned. */ + cmd_addr = adr & ~(map_bankwidth(map)-1); spin_lock(chip->mutex); ret = get_chip(map, chip, cmd_addr, FL_READY); @@ -1195,7 +1242,7 @@ static int cfi_intelext_read (struct mtd_info *mtd, loff_t from, size_t len, siz *retlen += thislen; len -= thislen; buf += thislen; - + ofs = 0; chipnum++; } @@ -1212,12 +1259,17 @@ static int __xipram do_write_oneword(struct map_info *map, struct flchip *chip, adr += chip->start; - /* Let's determine this according to the interleave only once */ + /* Let's determine those according to the interleave only once */ status_OK = CMD(0x80); switch (mode) { - case FL_WRITING: write_cmd = CMD(0x40); break; - case FL_OTP_WRITE: write_cmd = CMD(0xc0); break; - default: return -EINVAL; + case FL_WRITING: + write_cmd = (cfi->cfiq->P_ID != 0x0200) ? CMD(0x40) : CMD(0x41); + break; + case FL_OTP_WRITE: + write_cmd = CMD(0xc0); + break; + default: + return -EINVAL; } spin_lock(chip->mutex); @@ -1258,12 +1310,13 @@ static int __xipram do_write_oneword(struct map_info *map, struct flchip *chip, status = map_read(map, adr); if (map_word_andequal(map, status, status_OK, status_OK)) break; - + /* OK Still waiting */ if (time_after(jiffies, timeo)) { + map_write(map, CMD(0x70), adr); chip->state = FL_STATUS; xip_enable(map, chip, adr); - printk(KERN_ERR "waiting for chip to be ready timed out in word write\n"); + printk(KERN_ERR "%s: word write error (status timeout)\n", map->name); ret = -EIO; goto out; } @@ -1275,27 +1328,39 @@ static int __xipram do_write_oneword(struct map_info *map, struct flchip *chip, if (!z) { chip->word_write_time--; if (!chip->word_write_time) - chip->word_write_time++; + chip->word_write_time = 1; } - if (z > 1) + if (z > 1) chip->word_write_time++; /* Done and happy. */ chip->state = FL_STATUS; - /* check for lock bit */ - if (map_word_bitsset(map, status, CMD(0x02))) { - /* clear status */ + /* check for errors */ + if (map_word_bitsset(map, status, CMD(0x1a))) { + unsigned long chipstatus = MERGESTATUS(status); + + /* reset status */ map_write(map, CMD(0x50), adr); - /* put back into read status register mode */ map_write(map, CMD(0x70), adr); - ret = -EROFS; + xip_enable(map, chip, adr); + + if (chipstatus & 0x02) { + ret = -EROFS; + } else if (chipstatus & 0x08) { + printk(KERN_ERR "%s: word write error (bad VPP)\n", map->name); + ret = -EIO; + } else { + printk(KERN_ERR "%s: word write error (status 0x%lx)\n", map->name, chipstatus); + ret = -EINVAL; + } + + goto out; } xip_enable(map, chip, adr); out: put_chip(map, chip, adr); spin_unlock(chip->mutex); - return ret; } @@ -1328,7 +1393,7 @@ static int cfi_intelext_write_words (struct mtd_info *mtd, loff_t to , size_t le ret = do_write_oneword(map, &cfi->chips[chipnum], bus_ofs, datum, FL_WRITING); - if (ret) + if (ret) return ret; len -= n; @@ -1337,13 +1402,13 @@ static int cfi_intelext_write_words (struct mtd_info *mtd, loff_t to , size_t le (*retlen) += n; if (ofs >> cfi->chipshift) { - chipnum ++; + chipnum ++; ofs = 0; if (chipnum == cfi->numchips) return 0; } } - + while(len >= map_bankwidth(map)) { map_word datum = map_word_load(map, buf); @@ -1358,7 +1423,7 @@ static int cfi_intelext_write_words (struct mtd_info *mtd, loff_t to , size_t le len -= map_bankwidth(map); if (ofs >> cfi->chipshift) { - chipnum ++; + chipnum ++; ofs = 0; if (chipnum == cfi->numchips) return 0; @@ -1373,9 +1438,9 @@ static int cfi_intelext_write_words (struct mtd_info *mtd, loff_t to , size_t le ret = do_write_oneword(map, &cfi->chips[chipnum], ofs, datum, FL_WRITING); - if (ret) + if (ret) return ret; - + (*retlen) += len; } @@ -1383,20 +1448,24 @@ static int cfi_intelext_write_words (struct mtd_info *mtd, loff_t to , size_t le } -static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, - unsigned long adr, const u_char *buf, int len) +static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, + unsigned long adr, const struct kvec **pvec, + unsigned long *pvec_seek, int len) { struct cfi_private *cfi = map->fldrv_priv; - map_word status, status_OK; + map_word status, status_OK, write_cmd, datum; unsigned long cmd_adr, timeo; - int wbufsize, z, ret=0, bytes, words; + int wbufsize, z, ret=0, word_gap, words; + const struct kvec *vec; + unsigned long vec_seek; wbufsize = cfi_interleave(cfi) << cfi->cfiq->MaxBufWriteSize; adr += chip->start; cmd_adr = adr & ~(wbufsize-1); - + /* Let's determine this according to the interleave only once */ status_OK = CMD(0x80); + write_cmd = (cfi->cfiq->P_ID != 0x0200) ? CMD(0xe8) : CMD(0xe9); spin_lock(chip->mutex); ret = get_chip(map, chip, cmd_adr, FL_WRITING); @@ -1410,7 +1479,7 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, xip_disable(map, chip, cmd_adr); /* §4.8 of the 28FxxxJ3A datasheet says "Any time SR.4 and/or SR.5 is set - [...], the device will not accept any more Write to Buffer commands". + [...], the device will not accept any more Write to Buffer commands". So we must check here and reset those bits if they're set. Otherwise we're just pissing in the wind */ if (chip->state != FL_STATUS) @@ -1428,7 +1497,7 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, z = 0; for (;;) { - map_write(map, CMD(0xe8), cmd_adr); + map_write(map, write_cmd, cmd_adr); status = map_read(map, cmd_adr); if (map_word_andequal(map, status, status_OK, status_OK)) @@ -1446,41 +1515,66 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, map_write(map, CMD(0x50), cmd_adr); map_write(map, CMD(0x70), cmd_adr); xip_enable(map, chip, cmd_adr); - printk(KERN_ERR "Chip not ready for buffer write. status = %lx, Xstatus = %lx\n", - status.x[0], Xstatus.x[0]); + printk(KERN_ERR "%s: Chip not ready for buffer write. status = %lx, Xstatus = %lx\n", + map->name, status.x[0], Xstatus.x[0]); ret = -EIO; goto out; } } + /* Figure out the number of words to write */ + word_gap = (-adr & (map_bankwidth(map)-1)); + words = (len - word_gap + map_bankwidth(map) - 1) / map_bankwidth(map); + if (!word_gap) { + words--; + } else { + word_gap = map_bankwidth(map) - word_gap; + adr -= word_gap; + datum = map_word_ff(map); + } + /* Write length of data to come */ - bytes = len & (map_bankwidth(map)-1); - words = len / map_bankwidth(map); - map_write(map, CMD(words - !bytes), cmd_adr ); + map_write(map, CMD(words), cmd_adr ); /* Write data */ - z = 0; - while(z < words * map_bankwidth(map)) { - map_word datum = map_word_load(map, buf); - map_write(map, datum, adr+z); + vec = *pvec; + vec_seek = *pvec_seek; + do { + int n = map_bankwidth(map) - word_gap; + if (n > vec->iov_len - vec_seek) + n = vec->iov_len - vec_seek; + if (n > len) + n = len; - z += map_bankwidth(map); - buf += map_bankwidth(map); - } + if (!word_gap && len < map_bankwidth(map)) + datum = map_word_ff(map); - if (bytes) { - map_word datum; + datum = map_word_load_partial(map, datum, + vec->iov_base + vec_seek, + word_gap, n); - datum = map_word_ff(map); - datum = map_word_load_partial(map, datum, buf, 0, bytes); - map_write(map, datum, adr+z); - } + len -= n; + word_gap += n; + if (!len || word_gap == map_bankwidth(map)) { + map_write(map, datum, adr); + adr += map_bankwidth(map); + word_gap = 0; + } + + vec_seek += n; + if (vec_seek == vec->iov_len) { + vec++; + vec_seek = 0; + } + } while (len); + *pvec = vec; + *pvec_seek = vec_seek; /* GO GO GO */ map_write(map, CMD(0xd0), cmd_adr); chip->state = FL_WRITING; - INVALIDATE_CACHE_UDELAY(map, chip, + INVALIDATE_CACHE_UDELAY(map, chip, cmd_adr, len, chip->buffer_write_time); @@ -1506,13 +1600,14 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, /* OK Still waiting */ if (time_after(jiffies, timeo)) { + map_write(map, CMD(0x70), cmd_adr); chip->state = FL_STATUS; xip_enable(map, chip, cmd_adr); - printk(KERN_ERR "waiting for chip to be ready timed out in bufwrite\n"); + printk(KERN_ERR "%s: buffer write error (status timeout)\n", map->name); ret = -EIO; goto out; } - + /* Latency issues. Drop the lock, wait a while and retry */ z++; UDELAY(map, chip, cmd_adr, 1); @@ -1520,21 +1615,34 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, if (!z) { chip->buffer_write_time--; if (!chip->buffer_write_time) - chip->buffer_write_time++; + chip->buffer_write_time = 1; } - if (z > 1) + if (z > 1) chip->buffer_write_time++; /* Done and happy. */ chip->state = FL_STATUS; - /* check for lock bit */ - if (map_word_bitsset(map, status, CMD(0x02))) { - /* clear status */ + /* check for errors */ + if (map_word_bitsset(map, status, CMD(0x1a))) { + unsigned long chipstatus = MERGESTATUS(status); + + /* reset status */ map_write(map, CMD(0x50), cmd_adr); - /* put back into read status register mode */ - map_write(map, CMD(0x70), adr); - ret = -EROFS; + map_write(map, CMD(0x70), cmd_adr); + xip_enable(map, chip, cmd_adr); + + if (chipstatus & 0x02) { + ret = -EROFS; + } else if (chipstatus & 0x08) { + printk(KERN_ERR "%s: buffer write error (bad VPP)\n", map->name); + ret = -EIO; + } else { + printk(KERN_ERR "%s: buffer write error (status 0x%lx)\n", map->name, chipstatus); + ret = -EINVAL; + } + + goto out; } xip_enable(map, chip, cmd_adr); @@ -1543,70 +1651,65 @@ static int __xipram do_write_buffer(struct map_info *map, struct flchip *chip, return ret; } -static int cfi_intelext_write_buffers (struct mtd_info *mtd, loff_t to, - size_t len, size_t *retlen, const u_char *buf) +static int cfi_intelext_writev (struct mtd_info *mtd, const struct kvec *vecs, + unsigned long count, loff_t to, size_t *retlen) { struct map_info *map = mtd->priv; struct cfi_private *cfi = map->fldrv_priv; int wbufsize = cfi_interleave(cfi) << cfi->cfiq->MaxBufWriteSize; int ret = 0; int chipnum; - unsigned long ofs; + unsigned long ofs, vec_seek, i; + size_t len = 0; + + for (i = 0; i < count; i++) + len += vecs[i].iov_len; *retlen = 0; if (!len) return 0; chipnum = to >> cfi->chipshift; - ofs = to - (chipnum << cfi->chipshift); - - /* If it's not bus-aligned, do the first word write */ - if (ofs & (map_bankwidth(map)-1)) { - size_t local_len = (-ofs)&(map_bankwidth(map)-1); - if (local_len > len) - local_len = len; - ret = cfi_intelext_write_words(mtd, to, local_len, - retlen, buf); - if (ret) - return ret; - ofs += local_len; - buf += local_len; - len -= local_len; - - if (ofs >> cfi->chipshift) { - chipnum ++; - ofs = 0; - if (chipnum == cfi->numchips) - return 0; - } - } + ofs = to - (chipnum << cfi->chipshift); + vec_seek = 0; - while(len) { + do { /* We must not cross write block boundaries */ int size = wbufsize - (ofs & (wbufsize-1)); if (size > len) size = len; - ret = do_write_buffer(map, &cfi->chips[chipnum], - ofs, buf, size); + ret = do_write_buffer(map, &cfi->chips[chipnum], + ofs, &vecs, &vec_seek, size); if (ret) return ret; ofs += size; - buf += size; (*retlen) += size; len -= size; if (ofs >> cfi->chipshift) { - chipnum ++; + chipnum ++; ofs = 0; if (chipnum == cfi->numchips) return 0; } - } + } while (len); + return 0; } +static int cfi_intelext_write_buffers (struct mtd_info *mtd, loff_t to, + size_t len, size_t *retlen, const u_char *buf) +{ + struct kvec vec; + + vec.iov_base = (void *) buf; + vec.iov_len = len; + + return cfi_intelext_writev(mtd, &vec, 1, to, retlen); +} + static int __xipram do_erase_oneblock(struct map_info *map, struct flchip *chip, unsigned long adr, int len, void *thunk) { @@ -1672,23 +1775,17 @@ static int __xipram do_erase_oneblock(struct map_info *map, struct flchip *chip, status = map_read(map, adr); if (map_word_andequal(map, status, status_OK, status_OK)) break; - + /* OK Still waiting */ if (time_after(jiffies, timeo)) { - map_word Xstatus; map_write(map, CMD(0x70), adr); chip->state = FL_STATUS; - Xstatus = map_read(map, adr); - /* Clear status bits */ - map_write(map, CMD(0x50), adr); - map_write(map, CMD(0x70), adr); xip_enable(map, chip, adr); - printk(KERN_ERR "waiting for erase at %08lx to complete timed out. status = %lx, Xstatus = %lx.\n", - adr, status.x[0], Xstatus.x[0]); + printk(KERN_ERR "%s: block erase error: (status timeout)\n", map->name); ret = -EIO; goto out; } - + /* Latency issues. Drop the lock, wait a while and retry */ UDELAY(map, chip, adr, 1000000/HZ); } @@ -1698,43 +1795,40 @@ static int __xipram do_erase_oneblock(struct map_info *map, struct flchip *chip, chip->state = FL_STATUS; status = map_read(map, adr); - /* check for lock bit */ + /* check for errors */ if (map_word_bitsset(map, status, CMD(0x3a))) { - unsigned long chipstatus; + unsigned long chipstatus = MERGESTATUS(status); /* Reset the error bits */ map_write(map, CMD(0x50), adr); map_write(map, CMD(0x70), adr); xip_enable(map, chip, adr); - chipstatus = MERGESTATUS(status); - if ((chipstatus & 0x30) == 0x30) { - printk(KERN_NOTICE "Chip reports improper command sequence: status 0x%lx\n", chipstatus); - ret = -EIO; + printk(KERN_ERR "%s: block erase error: (bad command sequence, status 0x%lx)\n", map->name, chipstatus); + ret = -EINVAL; } else if (chipstatus & 0x02) { /* Protection bit set */ ret = -EROFS; } else if (chipstatus & 0x8) { /* Voltage */ - printk(KERN_WARNING "Chip reports voltage low on erase: status 0x%lx\n", chipstatus); + printk(KERN_ERR "%s: block erase error: (bad VPP)\n", map->name); ret = -EIO; - } else if (chipstatus & 0x20) { - if (retries--) { - printk(KERN_DEBUG "Chip erase failed at 0x%08lx: status 0x%lx. Retrying...\n", adr, chipstatus); - timeo = jiffies + HZ; - put_chip(map, chip, adr); - spin_unlock(chip->mutex); - goto retry; - } - printk(KERN_DEBUG "Chip erase failed at 0x%08lx: status 0x%lx\n", adr, chipstatus); + } else if (chipstatus & 0x20 && retries--) { + printk(KERN_DEBUG "block erase failed at 0x%08lx: status 0x%lx. Retrying...\n", adr, chipstatus); + timeo = jiffies + HZ; + put_chip(map, chip, adr); + spin_unlock(chip->mutex); + goto retry; + } else { + printk(KERN_ERR "%s: block erase failed at 0x%08lx (status 0x%lx)\n", map->name, adr, chipstatus); ret = -EIO; } - } else { - xip_enable(map, chip, adr); - ret = 0; + + goto out; } + xip_enable(map, chip, adr); out: put_chip(map, chip, adr); spin_unlock(chip->mutex); return ret; @@ -1754,7 +1848,7 @@ int cfi_intelext_erase_varsize(struct mtd_info *mtd, struct erase_info *instr) instr->state = MTD_ERASE_DONE; mtd_erase_callback(instr); - + return 0; } @@ -1775,7 +1869,7 @@ static void cfi_intelext_sync (struct mtd_info *mtd) if (!ret) { chip->oldstate = chip->state; chip->state = FL_SYNCING; - /* No need to wake_up() on this state change - + /* No need to wake_up() on this state change - * as the whole point is that nobody can do anything * with the chip now anyway. */ @@ -1789,7 +1883,7 @@ static void cfi_intelext_sync (struct mtd_info *mtd) chip = &cfi->chips[i]; spin_lock(chip->mutex); - + if (chip->state == FL_SYNCING) { chip->state = chip->oldstate; chip->oldstate = FL_READY; @@ -1846,7 +1940,7 @@ static int __xipram do_xxlock_oneblock(struct map_info *map, struct flchip *chip ENABLE_VPP(map); xip_disable(map, chip, adr); - + map_write(map, CMD(0x60), adr); if (thunk == DO_XXLOCK_ONEBLOCK_LOCK) { map_write(map, CMD(0x01), adr); @@ -1874,25 +1968,22 @@ static int __xipram do_xxlock_oneblock(struct map_info *map, struct flchip *chip status = map_read(map, adr); if (map_word_andequal(map, status, status_OK, status_OK)) break; - + /* OK Still waiting */ if (time_after(jiffies, timeo)) { - map_word Xstatus; map_write(map, CMD(0x70), adr); chip->state = FL_STATUS; - Xstatus = map_read(map, adr); xip_enable(map, chip, adr); - printk(KERN_ERR "waiting for unlock to complete timed out. status = %lx, Xstatus = %lx.\n", - status.x[0], Xstatus.x[0]); + printk(KERN_ERR "%s: block unlock error: (status timeout)\n", map->name); put_chip(map, chip, adr); spin_unlock(chip->mutex); return -EIO; } - + /* Latency issues. Drop the lock, wait a while and retry */ UDELAY(map, chip, adr, 1); } - + /* Done and happy. */ chip->state = FL_STATUS; xip_enable(map, chip, adr); @@ -1912,9 +2003,9 @@ static int cfi_intelext_lock(struct mtd_info *mtd, loff_t ofs, size_t len) ofs, len, 0); #endif - ret = cfi_varsize_frob(mtd, do_xxlock_oneblock, + ret = cfi_varsize_frob(mtd, do_xxlock_oneblock, ofs, len, DO_XXLOCK_ONEBLOCK_LOCK); - + #ifdef DEBUG_LOCK_BITS printk(KERN_DEBUG "%s: lock status after, ret=%d\n", __FUNCTION__, ret); @@ -1938,20 +2029,20 @@ static int cfi_intelext_unlock(struct mtd_info *mtd, loff_t ofs, size_t len) ret = cfi_varsize_frob(mtd, do_xxlock_oneblock, ofs, len, DO_XXLOCK_ONEBLOCK_UNLOCK); - + #ifdef DEBUG_LOCK_BITS printk(KERN_DEBUG "%s: lock status after, ret=%d\n", __FUNCTION__, ret); - cfi_varsize_frob(mtd, do_printlockstatus_oneblock, + cfi_varsize_frob(mtd, do_printlockstatus_oneblock, ofs, len, 0); #endif - + return ret; } #ifdef CONFIG_MTD_OTP -typedef int (*otp_op_t)(struct map_info *map, struct flchip *chip, +typedef int (*otp_op_t)(struct map_info *map, struct flchip *chip, u_long data_offset, u_char *buf, u_int size, u_long prot_offset, u_int groupno, u_int groupsize); @@ -2002,7 +2093,7 @@ do_otp_write(struct map_info *map, struct flchip *chip, u_long offset, datum = map_word_load_partial(map, datum, buf, gap, n); ret = do_write_oneword(map, chip, bus_ofs, datum, FL_OTP_WRITE); - if (ret) + if (ret) return ret; offset += n; @@ -2195,7 +2286,7 @@ static int cfi_intelext_lock_user_prot_reg(struct mtd_info *mtd, NULL, do_otp_lock, 1); } -static int cfi_intelext_get_fact_prot_info(struct mtd_info *mtd, +static int cfi_intelext_get_fact_prot_info(struct mtd_info *mtd, struct otp_info *buf, size_t len) { size_t retlen; @@ -2238,7 +2329,7 @@ static int cfi_intelext_suspend(struct mtd_info *mtd) if (chip->oldstate == FL_READY) { chip->oldstate = chip->state; chip->state = FL_PM_SUSPENDED; - /* No need to wake_up() on this state change - + /* No need to wake_up() on this state change - * as the whole point is that nobody can do anything * with the chip now anyway. */ @@ -2266,9 +2357,9 @@ static int cfi_intelext_suspend(struct mtd_info *mtd) if (ret) { for (i--; i >=0; i--) { chip = &cfi->chips[i]; - + spin_lock(chip->mutex); - + if (chip->state == FL_PM_SUSPENDED) { /* No need to force it into a known state here, because we're returning failure, and it didn't @@ -2279,8 +2370,8 @@ static int cfi_intelext_suspend(struct mtd_info *mtd) } spin_unlock(chip->mutex); } - } - + } + return ret; } @@ -2292,11 +2383,11 @@ static void cfi_intelext_resume(struct mtd_info *mtd) struct flchip *chip; for (i=0; i<cfi->numchips; i++) { - + chip = &cfi->chips[i]; spin_lock(chip->mutex); - + /* Go to known state. Chip may have been power cycled */ if (chip->state == FL_PM_SUSPENDED) { map_write(map, CMD(0xFF), cfi->chips[i].start); @@ -2318,7 +2409,7 @@ static int cfi_intelext_reset(struct mtd_info *mtd) struct flchip *chip = &cfi->chips[i]; /* force the completion of any ongoing operation - and switch to array mode so any bootloader in + and switch to array mode so any bootloader in flash is accessible for soft reboot. */ spin_lock(chip->mutex); ret = get_chip(map, chip, chip->start, FL_SYNCING); @@ -2355,20 +2446,23 @@ static void cfi_intelext_destroy(struct mtd_info *mtd) kfree(mtd->eraseregions); } -static char im_name_1[]="cfi_cmdset_0001"; -static char im_name_3[]="cfi_cmdset_0003"; +static char im_name_0001[] = "cfi_cmdset_0001"; +static char im_name_0003[] = "cfi_cmdset_0003"; +static char im_name_0200[] = "cfi_cmdset_0200"; static int __init cfi_intelext_init(void) { - inter_module_register(im_name_1, THIS_MODULE, &cfi_cmdset_0001); - inter_module_register(im_name_3, THIS_MODULE, &cfi_cmdset_0001); + inter_module_register(im_name_0001, THIS_MODULE, &cfi_cmdset_0001); + inter_module_register(im_name_0003, THIS_MODULE, &cfi_cmdset_0001); + inter_module_register(im_name_0200, THIS_MODULE, &cfi_cmdset_0001); return 0; } static void __exit cfi_intelext_exit(void) { - inter_module_unregister(im_name_1); - inter_module_unregister(im_name_3); + inter_module_unregister(im_name_0001); + inter_module_unregister(im_name_0003); + inter_module_unregister(im_name_0200); } module_init(cfi_intelext_init); |