From: Bean Huo <[email protected]>
Micron eMMC offers an additional set of commands that go beyond the JEDEC Device
Health Report. These additional DEVICE HEALTH commands are implemented using the
generic CMD56 command and they return a significant amount of useful information
about the status of the NAND device. Such as bad block counters, block erase
counters ,etc. For more information, refer to TN-FC-32: e·MMC Device Health Report.
Since the CMD56 is specified in JEDEC, to make CMD56 universal used and let more
users of mmc-utils get the benefit for this. I add CMD56 read, and let the user
to input the CMD56 argument, also, here I didn't add data parsing, just print raw
data, since it is vendor-specific.
Signed-off-by: Bean Huo <[email protected]>
---
mmc.c | 7 +++++++
mmc.h | 2 ++
mmc_cmds.c | 52 ++++++++++++++++++++++++++++++++++++++++++++++++++++
mmc_cmds.h | 1 +
4 files changed, 62 insertions(+)
diff --git a/mmc.c b/mmc.c
index eb2638b12271..ea465e59fdf3 100644
--- a/mmc.c
+++ b/mmc.c
@@ -237,6 +237,13 @@ static struct Command commands[] = {
"secure-trim1 | secure-trim2 | trim \n",
NULL
},
+ { do_general_cmd_read, -2,
+ "gen_cmd read", "<arg> <device>\n"
+ "Send GEN_CMD (CMD56) with specific <arg> to <device> to read vendor\n"
+ "specific data. <arg> must be 32 bits length hex number prefixed with 0x/0X.\n\n"
+ "NOTE!: Because this option is only used to read, the bit0 in <arg> must be 1",
+ NULL
+ },
{ 0, 0, 0, 0 }
};
diff --git a/mmc.h b/mmc.h
index 25d6864ac76f..b621374a1ed1 100644
--- a/mmc.h
+++ b/mmc.h
@@ -41,6 +41,8 @@
[1] Discard Enable
[0] Identify Write Blocks for
Erase (or TRIM Enable) R1b */
+#define MMC_GEN_CMD 56 /* adtc [31:1] stuff bits.
+ [0]: RD/WR1 R1 */
#define R1_OUT_OF_RANGE (1 << 31) /* er, c */
#define R1_ADDRESS_ERROR (1 << 30) /* erx, c */
diff --git a/mmc_cmds.c b/mmc_cmds.c
index 49d3e324d266..6e006b10d4fd 100644
--- a/mmc_cmds.c
+++ b/mmc_cmds.c
@@ -2981,3 +2981,55 @@ out:
return ret;
#endif
}
+
+int do_general_cmd_read(int nargs, char **argv)
+{
+ int dev_fd;
+ __u8 buf[512];
+ __u32 arg;
+ char *device;
+ char *endptr;
+ int ret = -EINVAL, i;
+ struct mmc_ioc_cmd idata;
+
+ if (nargs != 3) {
+ fprintf(stderr, "Usage: gen_cmd read <arg> </path/to/mmcblkX>\n");
+ exit(1);
+ }
+
+ device = argv[2];
+ dev_fd = open(device, O_RDWR);
+ if (dev_fd < 0) {
+ perror("device open failed");
+ exit(1);
+ }
+ arg = strtol(argv[1], &endptr, 16);
+ if (errno != 0 || *endptr != '\0' || !arg&0x1) {
+ fprintf(stderr, "Wrong ARG, it should be Hex number and bit0 must be 1\n");
+ goto out;
+ }
+ memset(&idata, 0, sizeof(idata));
+ idata.write_flag = 0;
+ idata.opcode = MMC_GEN_CMD;
+ idata.arg = arg;
+ idata.flags = MMC_RSP_SPI_R1 | MMC_RSP_R1 | MMC_CMD_ADTC;
+ idata.blksz = 512;
+ idata.blocks = 1;
+ mmc_ioc_cmd_set_data(idata, buf);
+
+ ret = ioctl(dev_fd, MMC_IOC_CMD, &idata);
+ if (ret) {
+ perror("ioctl");
+ goto out;
+ }
+
+ printf("Data:\n");
+ for (i = 0; i < 512; i++) {
+ printf("%2x ", buf[i]);
+ if ((i + 1) % 16 == 0)
+ printf("\n");
+ }
+out:
+ close(dev_fd);
+ return ret;
+}
diff --git a/mmc_cmds.h b/mmc_cmds.h
index 8331ab2373fd..0f7c0041f753 100644
--- a/mmc_cmds.h
+++ b/mmc_cmds.h
@@ -46,3 +46,4 @@ int do_read_scr(int argc, char **argv);
int do_read_cid(int argc, char **argv);
int do_read_csd(int argc, char **argv);
int do_erase(int nargs, char **argv);
+int do_general_cmd_read(int nargs, char **argv);
--
2.34.1
> On Sat, 2022-04-09 at 09:03 +0000, Avri Altman wrote:
> > > + memset(&idata, 0, sizeof(idata));
> > > + idata.write_flag = 0;
> > > + idata.opcode = MMC_GEN_CMD;
> > > + idata.arg = arg;
> > Practically arg is not needed, because bits[1..31] are meaningless.
> > Maybe just set: idata.arg = 1;
> >
> > Thanks,
> > Avri
> Thanks for your review.
>
> you are right arg is not needed according to eMMC spec. But as I mentioned
> in the commit message, for the universal use. we need it for the vendor-
> specific arg. I will change it that its arg will be 1 by default if the user doesn't
> specific arg.
Ack.
Thanks,
Avri
>
> Kind regards,
> Bean
>
>
On Sat, 2022-04-09 at 09:03 +0000, Avri Altman wrote:
> > + memset(&idata, 0, sizeof(idata));
> > + idata.write_flag = 0;
> > + idata.opcode = MMC_GEN_CMD;
> > + idata.arg = arg;
> Practically arg is not needed, because bits[1..31] are meaningless.
> Maybe just set: idata.arg = 1;
>
> Thanks,
> Avri
Thanks for your review.
you are right arg is not needed according to eMMC spec. But as I
mentioned in the commit message, for the universal use. we need it for
the vendor-specific arg. I will change it that its arg will be 1 by
default if the user doesn't specific arg.
Kind regards,
Bean
Tested-by: Rossler, Jakob (Nokia - DE/Ulm) <[email protected]>
Best regards
Jakob
-----Original Message-----
From: Avri Altman <[email protected]>
Sent: Monday, April 11, 2022 12:05 PM
To: Bean Huo <[email protected]>; [email protected]; [email protected]
Cc: [email protected]; [email protected]; Rossler, Jakob (Nokia - DE/Ulm) <[email protected]>; [email protected]; Bean Huo <[email protected]>
Subject: RE: [PATCH v1] mmc-utils: Add General command CMD56 read support
> On Sat, 2022-04-09 at 09:03 +0000, Avri Altman wrote:
> > > + memset(&idata, 0, sizeof(idata));
> > > + idata.write_flag = 0;
> > > + idata.opcode = MMC_GEN_CMD;
> > > + idata.arg = arg;
> > Practically arg is not needed, because bits[1..31] are meaningless.
> > Maybe just set: idata.arg = 1;
> >
> > Thanks,
> > Avri
> Thanks for your review.
>
> you are right arg is not needed according to eMMC spec. But as I
> mentioned in the commit message, for the universal use. we need it for
> the vendor- specific arg. I will change it that its arg will be 1 by
> default if the user doesn't specific arg.
Ack.
Thanks,
Avri
>
> Kind regards,
> Bean
>
>