Drop the IIO_ST macro for setting scan_type elements.
Subsystems (5)
Jonathan Cameron
maintainercover Drop the IIO_ST macro for setting scan_type elements.
This macro is no longer fit for purpose given it doesn't allow the endianness of the element to be set. Missuse of the macro has also been responsible for a couple of bugs including some fixed in this series.
If everyone is happy I'll be apply the fixes through the fixes-togreg branch and follow up with the others after those have worked their way through.
Please do check the fixes in particular as I'm not entirely sure they are correct (particularly the ad7887 which gave me a headache on a train last night at the end of a very long day).
Thanks,
Jonathan
Jonathan Cameron (24):
staging:iio:mag:hmc5843 fix incorrect endianness of channel as a
result of missuse of the IIO_ST macro.
iio:imu:adis16400 fix pressure channel scan type iio:adc:ad7887 Fix channel reported endianness from cpu to big endian iio:accel:bma180 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:mag:mag3110 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5064 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5360 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5380 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5421 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5446 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5449 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5504 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5624r replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5686 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5755 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5764 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:ad5791 replaces IIO_ST macro with explicit entries to struct
scan_type
iio:dac:max517 drop specification of scan type as unused in this
driver.
iio:dac:mcp4725 drop specification of scan type as unused in this
driver.
iio:light:adjd_s311 replaces IIO_ST macro with explicit entries to
struct scan_type
iio:light:tcs3472 replaces IIO_ST macro with explicit entries to
struct scan_type
staging:iio:ad7606 replaces IIO_ST macro with explicit entries to
struct scan_type
staging:iio:ad799x replaces IIO_ST macro with explicit entries to
struct scan_type
iio:drop IIO_ST macro
drivers/iio/accel/bma180.c | 7 ++++++- drivers/iio/adc/ad7887.c | 16 ++++++++++++++-- drivers/iio/dac/ad5064.c | 7 ++++++- drivers/iio/dac/ad5360.c | 7 ++++++- drivers/iio/dac/ad5380.c | 7 ++++++- drivers/iio/dac/ad5421.c | 6 +++++- drivers/iio/dac/ad5446.c | 9 +++++++-- drivers/iio/dac/ad5449.c | 7 ++++++- drivers/iio/dac/ad5504.c | 6 +++++- drivers/iio/dac/ad5624r_spi.c | 7 ++++++- drivers/iio/dac/ad5686.c | 9 +++++++-- drivers/iio/dac/ad5755.c | 7 ++++++- drivers/iio/dac/ad5764.c | 7 ++++++- drivers/iio/dac/ad5791.c | 9 +++++++-- drivers/iio/dac/max517.c | 1 - drivers/iio/dac/mcp4725.c | 1 - drivers/iio/imu/adis16400_core.c | 7 ++++++- drivers/iio/light/adjd_s311.c | 7 ++++++- drivers/iio/light/tcs3472.c | 7 ++++++- drivers/iio/magnetometer/mag3110.c | 6 +++++- drivers/staging/iio/adc/ad7606_core.c | 7 ++++++- drivers/staging/iio/adc/ad799x_core.c | 8 +++++++- drivers/staging/iio/magnetometer/hmc5843.c | 7 ++++++- include/linux/iio/iio.h | 16 ++++++++++------ 24 files changed, 145 insertions(+), 33 deletions(-)
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 1/24 staging:iio:mag:hmc5843 fix incorrect endianness of channel as a result of missuse of the IIO_ST macro.
This driver sets the shift value equal to IIO_BE (or 1) rather than setting that to 0 and specificying the endianness. This means the channel type is missreported as
[be|le]:u16/16>>1 where the be|le is dependent on the cpu native endianness, rather than
be:u16/16>>0 resulting in any userspace code using this information, miss converting the channel and generating thoroughly trashed data.
--- drivers/staging/iio/magnetometer/hmc5843.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/staging/iio/magnetometer/hmc5843.c +6 -1
@@ -451,7 +451,12 @@ done: .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) | \ BIT(IIO_CHAN_INFO_SAMP_FREQ), \ .scan_index = idx, \ - .scan_type = IIO_ST('s', 16, 16, IIO_BE), \ + .scan_type = { \ + .sign = 's', \ + .realbits = 16, \ + .storagebits = 16, \ + .endianness = IIO_BE, \ + }, \ }static const struct iio_chan_spec hmc5843_channels[] = {
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerRe: [PATCH 01/24] staging:iio:mag:hmc5843 fix incorrect endianness of channel as a result of missuse of the IIO_ST macro.
Quoted 9 lines · Jonathan Cameron
This driver sets the shift value equal to
IIO_BE(or 1) rather than setting that to 0 and specificying the endianness. This means the channel type is missreported as
[be|le]:u16/16>>1 where the be|le is dependent on the cpu native endianness, rather than
be:u16/16>>0 resulting in any userspace code using this information, miss converting the channel and generating thoroughly trashed data.Signed-off-by Jonathan Cameron <[email protected]>
Applied with the formatting fixup to the fixes-togreg branch of iio.git.
Tagged for stable.
Quoted 23 lines
--- drivers/staging/iio/magnetometer/hmc5843.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)drivers/staging/iio/magnetometer/hmc5843.c +6
-1@@ -451,7 +451,12 @@ done: .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) | \ BIT(IIO_CHAN_INFO_SAMP_FREQ), \ .scan_index = idx, \ - .scan_type = IIO_ST('s', 16, 16, IIO_BE), \ + .scan_type = { \ + .sign = 's', \ + .realbits = 16, \ + .storagebits = 16, \ + .endianness = IIO_BE, \ + }, \ }static const struct iio_chan_spec hmc5843_channels[] = {
Jonathan Cameron
maintainerPATCH 2/24 iio:imu:adis16400 fix pressure channel scan type
In reply to Jonathan CameronA single channel in this driver was using the IIO_ST macro. This does not provide a parameter for setting the endianness of the channel. Thus this channel will have been reported as whatever is the native endianness of the cpu rather than big endian. This means it would be incorrect on little endian platforms.
--- drivers/iio/imu/adis16400_core.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/imu/adis16400_core.c +6 -1
@@ -651,7 +651,12 @@ static const struct iio_chan_spec adis16448_channels[] = { .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .address = ADIS16448_BARO_OUT, .scan_index = ADIS16400_SCAN_BARO, - .scan_type =IIO_ST ('s' ,16 ,16 ,0 ), + .scan_type = { + .sign = 's', + .realbits = 16, + .storagebits = 16, + .endianness =IIO_BE , + }, },ADIS16400_TEMP_CHAN (ADIS16448_TEMP_OUT ,12 ),IIO_CHAN_SOFT_TIMESTAMP (11 )
Signature · 1 line
1.8.4.2
Re: [PATCH 02/24] iio:imu:adis16400 fix pressure channel scan type
Quoted 7 lines · Jonathan Cameron
A single channel in this driver was using the
IIO_STmacro. This does not provide a parameter for setting the endianness of the channel. Thus this channel will have been reported as whatever is the native endianness of the cpu rather than big endian. This means it would be incorrect on little endian platforms.Signed-off-by Jonathan Cameron <[email protected]>
Applied to the fixes-togreg branch of iio.git
Quoted 23 lines
--- drivers/iio/imu/adis16400_core.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)drivers/iio/imu/adis16400_core.c +6
-1@@ -651,7 +651,12 @@ static const struct iio_chan_spec adis16448_channels[] = { .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .address = ADIS16448_BARO_OUT, .scan_index = ADIS16400_SCAN_BARO, - .scan_type =IIO_ST ('s' ,16 ,16 ,0 ), + .scan_type = { + .sign = 's', + .realbits = 16, + .storagebits = 16, + .endianness =IIO_BE , + }, },ADIS16400_TEMP_CHAN (ADIS16448_TEMP_OUT ,12 ),IIO_CHAN_SOFT_TIMESTAMP (11 )
Jonathan Cameron
maintainerPATCH 3/24 iio:adc:ad7887 Fix channel reported endianness from cpu to big endian
In reply to Jonathan CameronNote this also sets the endianness to big endian whereas it would previously have defaulted to the cpu endian. Hence technically this is a bug fix on LE platforms.
--- drivers/iio/adc/ad7887.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)
drivers/iio/adc/ad7887.c +14 -2
@@ -200,7 +200,13 @@ static const struct ad7887_chip_info ad7887_chip_info_tbl[] = { .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .address = 1, .scan_index = 1, - .scan_type = IIO_ST('u', 12, 16, 0), + .scan_type = { + .sign = 'u', + .realbits = 12, + .storagebits = 16, + .shift = 0, + .endianness =IIO_BE , + }, }, .channel[1] = { .type = IIO_VOLTAGE,
@@ -210,7 +216,13 @@ static const struct ad7887_chip_info ad7887_chip_info_tbl[] = { .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .address = 0, .scan_index = 0, - .scan_type = IIO_ST('u', 12, 16, 0), + .scan_type = { + .sign = 'u', + .realbits = 12, + .storagebits = 16, + .shift = 0, + .endianness =IIO_BE , + }, }, .channel[2] = IIO_CHAN_SOFT_TIMESTAMP(2), .int_vref_mv = 2500,
Signature · 1 line
1.8.4.2
Re: [PATCH 03/24] iio:adc:ad7887 Fix channel reported endianness from cpu to big endian
Quoted 5 lines · Jonathan Cameron
Note this also sets the endianness to big endian whereas it would previously have defaulted to the cpu endian. Hence technically this is a bug fix on LE platforms.
Signed-off-by Jonathan Cameron <[email protected]>
Applied to the fixes togreg branch of iio.git with the formatting fix.
Quoted 39 lines
--- drivers/iio/adc/ad7887.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)drivers/iio/adc/ad7887.c +14
-2@@ -200,7 +200,13 @@ static const struct ad7887_chip_info ad7887_chip_info_tbl[] = { .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .address = 1, .scan_index = 1, - .scan_type = IIO_ST('u', 12, 16, 0), + .scan_type = { + .sign = 'u', + .realbits = 12, + .storagebits = 16, + .shift = 0, + .endianness =IIO_BE , + }, }, .channel[1] = { .type = IIO_VOLTAGE,@@ -210,7 +216,13 @@ static const struct ad7887_chip_info ad7887_chip_info_tbl[] = { .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), .address = 0, .scan_index = 0, - .scan_type = IIO_ST('u', 12, 16, 0), + .scan_type = { + .sign = 'u', + .realbits = 12, + .storagebits = 16, + .shift = 0, + .endianness =IIO_BE , + }, }, .channel[2] = IIO_CHAN_SOFT_TIMESTAMP(2), .int_vref_mv = 2500,
Jonathan Cameron
maintainerPATCH 4/24 iio:accel:bma180 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronThe IIO_ST macro no longer covers all the elements of struct scan_type and has this has lead to some bugs being introduced.
The drivers are easier to follow with this structure being directly filled so that is now preferred.
--- drivers/iio/accel/bma180.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/accel/bma180.c +6 -1
@@ -455,7 +455,12 @@ static const struct iio_chan_spec_ext_info bma180_ext_info[] = {BIT (IIO_CHAN_INFO_LOW_PASS_FILTER_3DB_FREQUENCY ), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ .scan_index = (_index), \ - .scan_type = IIO_ST('s', 14, 16, 2), \ + .scan_type = { \ + .sign ='s' , \ + .realbits =14 , \ + .storagebits =16 , \ + .shift =2 , \ + }, \ .ext_info = bma180_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 5/24 iio:mag:mag3110 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronThe IIO_ST macro no longer covers all the elements of struct scan_type and has this has lead to some bugs being introduced.
The drivers are easier to follow with this structure being directly filled so that is now preferred.
--- drivers/iio/magnetometer/mag3110.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/iio/magnetometer/mag3110.c +5 -1
@@ -266,7 +266,11 @@ static const struct iio_chan_spec mag3110_channels[] = { .type = IIO_TEMP, .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), .scan_index = 3, - .scan_type = IIO_ST('s', 8, 8, 0), + .scan_type = { + .sign = 's', + .realbits = 8, + .storagebits = 8, + }, }, IIO_CHAN_SOFT_TIMESTAMP(4), };
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 6/24 iio:dac:ad5064 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5064.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5064.c +6 -1
@@ -299,7 +299,12 @@ static const struct iio_chan_spec_ext_info ad5064_ext_info[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) | \ BIT(IIO_CHAN_INFO_SCALE), \ .address = addr, \ - .scan_type = IIO_ST('u', (bits), 16, 20 - (bits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (bits), \ + .storagebits = 16, \ + .shift = 20 - bits, \ + }, \ .ext_info = ad5064_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 7/24 iio:dac:ad5360 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5360.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5360.c +6 -1
@@ -107,7 +107,12 @@ enum ad5360_type { BIT(IIO_CHAN_INFO_OFFSET) | \ BIT(IIO_CHAN_INFO_CALIBSCALE) | \ BIT(IIO_CHAN_INFO_CALIBBIAS), \ - .scan_type = IIO_ST('u', (bits), 16, 16 - (bits)) \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (bits), \ + .storagebits = 16, \ + .shift = 16 - (bits), \ + }, \ }static const struct ad5360_chip_info ad5360_chip_info_tbl[] = {
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 8/24 iio:dac:ad5380 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5380.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5380.c +6 -1
@@ -261,7 +261,12 @@ static struct iio_chan_spec_ext_info ad5380_ext_info[] = {BIT (IIO_CHAN_INFO_CALIBSCALE ) | \ BIT(IIO_CHAN_INFO_CALIBBIAS), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ - .scan_type = IIO_ST('u', (_bits), 16, 14 - (_bits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (_bits), \ + .storagebits = 16, \ + .shift = 14 - (_bits), \ + }, \ .ext_info = ad5380_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 9/24 iio:dac:ad5421 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5421.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5421.c +5 -1
@@ -114,7 +114,11 @@ static const struct iio_chan_spec ad5421_channels[] = {BIT (IIO_CHAN_INFO_CALIBBIAS ), .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) |BIT (IIO_CHAN_INFO_OFFSET ), - .scan_type = IIO_ST('u', 16, 16, 0), + .scan_type = { + .sign ='u' , + .realbits =16 , + .storagebits =16 , + }, .event_spec = ad5421_current_event, .num_event_specs = ARRAY_SIZE(ad5421_current_event), },
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 10/24 iio:dac:ad5446 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5446.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/iio/dac/ad5446.c +7 -2
@@ -139,14 +139,19 @@ static const struct iio_chan_spec_ext_info ad5446_ext_info_powerdown[] = { { }, }; -#define _AD5446_CHANNEL (bits, storage, shift, ext) { \ +#define _AD5446_CHANNEL (bits, storage, _shift, ext) { \ .type = IIO_VOLTAGE, \ .indexed = 1, \ .output = 1, \ .channel = 0, \ .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ - .scan_type = IIO_ST('u', (bits), (storage), (shift)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (bits), \ + .storagebits = (storage), \ + .shift = (_shift), \ + }, \ .ext_info = (ext), \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 11/24 iio:dac:ad5449 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5449.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5449.c +6 -1
@@ -204,7 +204,12 @@ static const struct iio_info ad5449_info = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) | \ BIT(IIO_CHAN_INFO_SCALE), \ .address = (chan), \ - .scan_type = IIO_ST('u', (bits), 16, 12 - (bits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (bits), \ + .storagebits = 16, \ + .shift = 12 - (bits), \ + }, \ }#define DECLARE_AD5449_CHANNELS (name, bits) \
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 12/24 iio:dac:ad5504 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5504.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5504.c +5 -1
@@ -261,7 +261,11 @@ static const struct iio_chan_spec_ext_info ad5504_ext_info[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ .address = AD5504_ADDR_DAC(_chan), \ - .scan_type = IIO_ST('u', 12, 16, 0), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = 12, \ + .storagebits = 16, \ + }, \ .ext_info = ad5504_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 13/24 iio:dac:ad5624r replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5624r_spi.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5624r_spi.c +6 -1
@@ -176,7 +176,12 @@ static const struct iio_chan_spec_ext_info ad5624r_ext_info[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ .address = (_chan), \ - .scan_type = IIO_ST('u', (_bits), 16, 16 - (_bits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (_bits), \ + .storagebits = 16, \ + .shift = 16 - (_bits), \ + }, \ .ext_info = ad5624r_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 14/24 iio:dac:ad5686 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5686.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/iio/dac/ad5686.c +7 -2
@@ -267,7 +267,7 @@ static const struct iio_chan_spec_ext_info ad5686_ext_info[] = { { }, }; -#define AD5868_CHANNEL (chan, bits, shift) { \ +#define AD5868_CHANNEL (chan, bits, _shift) { \ .type = IIO_VOLTAGE, \ .indexed = 1, \ .output = 1, \
@@ -275,7 +275,12 @@ static const struct iio_chan_spec_ext_info ad5686_ext_info[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE),\ .address = AD5686_ADDR_DAC(chan), \ - .scan_type = IIO_ST('u', bits, 16, shift), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (bits), \ + .storagebits = 16, \ + .shift = (_shift), \ + }, \ .ext_info = ad5686_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 15/24 iio:dac:ad5755 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5755.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5755.c +6 -1
@@ -392,7 +392,12 @@ static const struct iio_chan_spec_ext_info ad5755_ext_info[] = {BIT (IIO_CHAN_INFO_OFFSET ) | \BIT (IIO_CHAN_INFO_CALIBSCALE ) | \ BIT(IIO_CHAN_INFO_CALIBBIAS), \ - .scan_type = IIO_ST('u', (_bits), 16, 16 - (_bits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (_bits), \ + .storagebits = 16, \ + .shift = 16 - (_bits), \ + }, \ .ext_info = ad5755_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 16/24 iio:dac:ad5764 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5764.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/dac/ad5764.c +6 -1
@@ -83,7 +83,12 @@ enum ad5764_type {BIT (IIO_CHAN_INFO_CALIBSCALE ) | \ BIT(IIO_CHAN_INFO_CALIBBIAS), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_OFFSET), \ - .scan_type = IIO_ST('u', (_bits), 16, 16 - (_bits)) \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (_bits), \ + .storagebits = 16, \ + .shift =16 - (_bits), \ + }, \ } #define DECLARE_AD5764_CHANNELS(_name, _bits) \
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 17/24 iio:dac:ad5791 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/dac/ad5791.c | 9 +++++++-- 1 file changed, 7 insertions(+), 2 deletions(-)
drivers/iio/dac/ad5791.c +7 -2
@@ -293,7 +293,7 @@ static const struct iio_chan_spec_ext_info ad5791_ext_info[] = { { }, }; -#define AD5791_CHAN (bits, shift) { \ +#define AD5791_CHAN (bits, _shift) { \ .type = IIO_VOLTAGE, \ .output = 1, \ .indexed = 1, \
@@ -302,7 +302,12 @@ static const struct iio_chan_spec_ext_info ad5791_ext_info[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE) | \ BIT(IIO_CHAN_INFO_OFFSET), \ - .scan_type = IIO_ST('u', bits, 24, shift), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (bits), \ + .storagebits = 24, \ + .shift = (_shift), \ + }, \ .ext_info = ad5791_ext_info, \ }
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 18/24 iio:dac:max517 drop specification of scan type as unused in this driver.
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain so the simplest path with this driver is to not specify the unused scan type.
--- drivers/iio/dac/max517.c | 1 - 1 file changed, 1 deletion(-)
drivers/iio/dac/max517.c +0 -1
@@ -146,7 +146,6 @@ static const struct iio_info max517_info = { .channel = (chan), \ .info_mask_separate = BIT(IIO_CHAN_INFO_RAW) | \ BIT(IIO_CHAN_INFO_SCALE), \ - .scan_type = IIO_ST('u', 8, 8, 0), \ }static const struct iio_chan_spec max517_channels[] = {
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 19/24 iio:dac:mcp4725 drop specification of scan type as unused in this driver.
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain so the simplest path with this driver is to not specify the unused scan type.
--- drivers/iio/dac/mcp4725.c | 1 - 1 file changed, 1 deletion(-)
drivers/iio/dac/mcp4725.c +0 -1
@@ -209,7 +209,6 @@ static const struct iio_chan_spec mcp4725_channel = { .channel = 0, .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), - .scan_type = IIO_ST('u', 12, 16, 0), .ext_info = mcp4725_ext_info, };
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 20/24 iio:light:adjd_s311 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/light/adjd_s311.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/light/adjd_s311.c +6 -1
@@ -155,7 +155,12 @@ done:BIT (IIO_CHAN_INFO_INT_TIME ), \ .channel2 = (IIO_MOD_LIGHT_##_color), \ .scan_index = (_scan_idx), \ - .scan_type = IIO_ST('u', 10, 16, 0), \ + .scan_type = { \ + .sign ='u' , \ + .realbits =10 , \ + .storagebits =16 , \ + .endianness =IIO_CPU , \ + }, \ } static const struct iio_chan_spec adjd_s311_channels[] = {
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 21/24 iio:light:tcs3472 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain.
--- drivers/iio/light/tcs3472.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/iio/light/tcs3472.c +6 -1
@@ -67,7 +67,12 @@ struct tcs3472_data { .channel2 = IIO_MOD_LIGHT_##_color, \ .address = _addr, \ .scan_index = _si, \ - .scan_type = IIO_ST('u', 16, 16, 0), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = 16, \ + .storagebits = 16, \ + .endianness = IIO_CPU, \ + }, \ }static const int tcs3472_agains[] = {1 ,4 ,16 ,60 };
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 22/24 staging:iio:ad7606 replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain. Note that this driver currently ensures all fields are converted to the native endianness to keep things consistent across the different bus types.
--- drivers/staging/iio/adc/ad7606_core.c | 7 ++++++- 1 file changed, 6 insertions(+), 1 deletion(-)
drivers/staging/iio/adc/ad7606_core.c +6 -1
@@ -239,7 +239,12 @@ static const struct attribute_group ad7606_attribute_group_range = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE),\ .scan_index = num, \ - .scan_type = IIO_ST('s', 16, 16, 0), \ + .scan_type = { \ + .sign = 's', \ + .realbits = 16, \ + .storagebits = 16, \ + .endianness = IIO_CPU, \ + }, \ }static const struct iio_chan_spec ad7606_8_channels[] = {
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerPATCH 23/24 staging:iio:ad799x replaces IIO_ST macro with explicit entries to struct scan_type
In reply to Jonathan CameronIIO_ST is going away as it is a pain to maintain. As an i2c driver, multi byte words will be read by the driver in the native cpu endianness.
--- drivers/staging/iio/adc/ad799x_core.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
drivers/staging/iio/adc/ad799x_core.c +7 -1
@@ -409,7 +409,13 @@ static const struct iio_event_spec ad799x_events[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ .scan_index = (_index), \ - .scan_type = IIO_ST('u', _realbits, 16, 12 - (_realbits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (_realbits), \ + .storagebits = 16, \ + .shift = 12 - (_realbits), \ + .endianness =IIO_CPU , \ + }, \ .event_spec = _ev_spec, \ .num_event_specs = _num_ev_spec, \ }
Signature · 1 line
1.8.4.2
Re: [PATCH 23/24] staging:iio:ad799x replaces IIO_ST macro with explicit entries to struct scan_type
Quoted 2 lines · Jonathan Cameron
IIO_STis going away as it is a pain to maintain. As an i2c driver, multi byte words will be read by the driver in the native cpu endianness.
Are you sure about this? As far as I understand it the I2C message will be read byte by byte. The device sends the MSB first, so I think it should be IIO_BE.
Quoted 26 lines
Signed-off-by Jonathan Cameron <[email protected]>--- drivers/staging/iio/adc/ad799x_core.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/staging/iio/adc/ad799x_core.c +7
-1@@ -409,7 +409,13 @@ static const struct iio_event_spec ad799x_events[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ .scan_index = (_index), \ - .scan_type = IIO_ST('u', _realbits, 16, 12 - (_realbits)), \ + .scan_type = { \ + .sign = 'u', \ + .realbits = (_realbits), \ + .storagebits = 16, \ + .shift = 12 - (_realbits), \ + .endianness =IIO_CPU , \ + }, \ .event_spec = _ev_spec, \ .num_event_specs = _num_ev_spec, \ }
Jonathan Cameron
maintainerRe: [PATCH 23/24] staging:iio:ad799x replaces IIO_ST macro with explicit entries to struct scan_type
Quoted 7 lines · Lars-Peter Clausen
Quoted 2 lines · Jonathan Cameron
IIO_STis going away as it is a pain to maintain. As an i2c driver, multi byte words will be read by the driver in the native cpu endianness.Are you sure about this? As far as I understand it the I2C message will be read byte by byte. The device sends the MSB first, so I think it should be
IIO_BE.
Gah. I just spent 5 minutes writing an essay on the fact it was an smbus word read call, only to realise I was looking at the wrong function and this is done with a block read. Doh. You are quite right! I'll switch this one to IIO_BE.
Quoted 33 lines
Quoted 17 lines
Signed-off-by Jonathan Cameron <[email protected]>--- drivers/staging/iio/adc/ad799x_core.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)drivers/staging/iio/adc/ad799x_core.c +0
-1@@ -409,7 +409,13 @@ static const struct iio_event_spec ad799x_events[] = { .info_mask_separate = BIT(IIO_CHAN_INFO_RAW), \ .info_mask_shared_by_type = BIT(IIO_CHAN_INFO_SCALE), \ .scan_index = (_index), \ - .scan_type = IIO_ST('u', _realbits, 16, 12 - (_realbits)), \drivers/staging/iio/adc/ad799x_core.c static const struct iio_event_spec ad799x_events[] = { · line 421
Quoted 10 lines
drivers/staging/iio/adc/ad799x_core.c +7
-0+ .scan_type = { \ + .sign = 'u', \ + .realbits = (_realbits), \ + .storagebits = 16, \ + .shift = 12 - (_realbits), \ + .endianness =IIO_CPU , \ + }, \ .event_spec = _ev_spec, \ .num_event_specs = _num_ev_spec, \ }--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Jonathan Cameron
maintainerPATCH 24/24 iio:drop IIO_ST macro
In reply to Jonathan CameronThis macro no longer allows all the elements of the scan_type structure to be set. Missinterpretation of the parameters also caused a couple of recent bugs. No mainline drivers now use this macro so drop it.
--- include/linux/iio/iio.h | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-)
include/linux/iio/iio.h +10 -6
@@ -254,12 +254,16 @@ static inline bool iio_channel_has_info(const struct iio_chan_spec *chan, (chan->info_mask_shared_by_all &BIT (type)); } -#define IIO_ST (si, rb, sb, sh) \ - { .sign = si, .realbits = rb, .storagebits = sb, .shift = sh } - -#define IIO_CHAN_SOFT_TIMESTAMP (_si) \ - { .type = IIO_TIMESTAMP, .channel = -1, \ - .scan_index = _si, .scan_type = IIO_ST('s', 64, 64, 0) } +#define IIO_CHAN_SOFT_TIMESTAMP (_si) { \ + .type = IIO_TIMESTAMP, \ + .channel = -1, \ + .scan_index = _si, \ + .scan_type = { \ + .sign = 's', \ + .realbits = 64, \ + .storagebits = 64, \ + }, \ +}/** * iio_get_time_ns() - utility function to get a time stamp for events etc
Signature · 1 line
1.8.4.2
Jonathan Cameron
maintainerRe: [PATCH 24/24] iio:drop IIO_ST macro
Quoted 5 lines · Jonathan Cameron
This macro no longer allows all the elements of the
scan_typestructure to be set. Missinterpretation of the parameters also caused a couple of recent bugs. No mainline drivers now use this macro so drop it.Signed-off-by Jonathan Cameron <[email protected]>
Finally applied to the togreg branch of iio.git. It's gone at last!
J
Quoted 32 lines
--- include/linux/iio/iio.h | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-)include/linux/iio/iio.h +10
-6@@ -254,12 +254,16 @@ static inline bool iio_channel_has_info(const struct iio_chan_spec *chan, (chan->info_mask_shared_by_all &BIT (type)); } -#define IIO_ST (si, rb, sb, sh) \ - { .sign = si, .realbits = rb, .storagebits = sb, .shift = sh } - -#define IIO_CHAN_SOFT_TIMESTAMP (_si) \ - { .type = IIO_TIMESTAMP, .channel = -1, \ - .scan_index = _si, .scan_type = IIO_ST('s', 64, 64, 0) } +#define IIO_CHAN_SOFT_TIMESTAMP (_si) { \ + .type = IIO_TIMESTAMP, \ + .channel = -1, \ + .scan_index = _si, \ + .scan_type = { \ + .sign = 's', \ + .realbits = 64, \ + .storagebits = 64, \ + }, \ +}/** * iio_get_time_ns() - utility function to get a time stamp for events etc
Quoted 11 lines · Jonathan Cameron
This macro is no longer fit for purpose given it doesn't allow the endianness of the element to be set. Missuse of the macro has also been responsible for a couple of bugs including some fixed in this series.
If everyone is happy I'll be apply the fixes through the fixes-togreg branch and follow up with the others after those have worked their way through.
Please do check the fixes in particular as I'm not entirely sure they are correct (particularly the ad7887 which gave me a headache on a train last night at the end of a very long day).
Looks all good to me, except for the ad799x.
One minor nitpick though. I think the proper indention is
{
.foo = bar,
}instead of
{
.foo = bar,
}- Lars
Quoted 77 lines
Thanks,
Jonathan
Jonathan Cameron (24):
staging:iio:mag:hmc5843 fix incorrect endianness of channel as a
result of missuse of theIIO_STmacro.
iio:imu:adis16400 fix pressure channel scan type iio:adc:ad7887 Fix channel reported endianness from cpu to big endian iio:accel:bma180 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:mag:mag3110 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5064 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5360 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5380 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5421 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5446 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5449 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5504 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5624r replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5686 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5755 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5764 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5791 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:max517 drop specification of scan type as unused in this
driver.
iio:dac:mcp4725 drop specification of scan type as unused in this
driver.
iio:light:adjd_s311 replacesIIO_STmacro with explicit entries to
structscan_type
iio:light:tcs3472 replacesIIO_STmacro with explicit entries to
structscan_type
staging:iio:ad7606 replacesIIO_STmacro with explicit entries to
structscan_type
staging:iio:ad799x replacesIIO_STmacro with explicit entries to
structscan_type
iio:dropIIO_STmacrodrivers/iio/accel/bma180.c | 7 ++++++- drivers/iio/adc/ad7887.c | 16 ++++++++++++++-- drivers/iio/dac/ad5064.c | 7 ++++++- drivers/iio/dac/ad5360.c | 7 ++++++- drivers/iio/dac/ad5380.c | 7 ++++++- drivers/iio/dac/ad5421.c | 6 +++++- drivers/iio/dac/ad5446.c | 9 +++++++-- drivers/iio/dac/ad5449.c | 7 ++++++- drivers/iio/dac/ad5504.c | 6 +++++- drivers/iio/dac/ad5624r_spi.c | 7 ++++++- drivers/iio/dac/ad5686.c | 9 +++++++-- drivers/iio/dac/ad5755.c | 7 ++++++- drivers/iio/dac/ad5764.c | 7 ++++++- drivers/iio/dac/ad5791.c | 9 +++++++-- drivers/iio/dac/max517.c | 1 - drivers/iio/dac/mcp4725.c | 1 - drivers/iio/imu/adis16400_core.c | 7 ++++++- drivers/iio/light/adjd_s311.c | 7 ++++++- drivers/iio/light/tcs3472.c | 7 ++++++- drivers/iio/magnetometer/mag3110.c | 6 +++++- drivers/staging/iio/adc/ad7606_core.c | 7 ++++++- drivers/staging/iio/adc/ad799x_core.c | 8 +++++++- drivers/staging/iio/magnetometer/hmc5843.c | 7 ++++++- include/linux/iio/iio.h | 16 ++++++++++------ 24 files changed, 145 insertions(+), 33 deletions(-)
Jonathan Cameron
maintainerQuoted 8 lines · Lars-Peter Clausen
Quoted 7 lines · Jonathan Cameron
This macro is no longer fit for purpose given it doesn't allow the endianness of the element to be set. Missuse of the macro has also been responsible for a couple of bugs including some fixed in this series.
If everyone is happy I'll be apply the fixes through the fixes-togreg branch and follow up with the others after those have worked their way through.
Quoted 8 lines
Quoted 4 lines
Please do check the fixes in particular as I'm not entirely sure they are correct (particularly the ad7887 which gave me a headache on a train last night at the end of a very long day).
Looks all good to me, except for the ad799x.
Thanks. I've applied 1..3 to fixes-togreg and 4..22 to the togreg branch of iio.git (pushed out as testing for build tests).
If I'd been more awake I'd have noticed that with the ad799x now being IIO_BE that one is another fix. Hence I've applied 23 with a new blurb to fixes-togreg.
Will apply 24 once all the others in are in one tree!
Indentation fixed up as you pointed out below.
Jonathan
Quoted 97 lines
One minor nitpick though. I think the proper indention is
{ .foo = bar, }instead of
{ .foo = bar, }- Lars
Quoted 68 lines
Thanks,
Jonathan
Jonathan Cameron (24):
staging:iio:mag:hmc5843 fix incorrect endianness of channel as a
result of missuse of theIIO_STmacro.
iio:imu:adis16400 fix pressure channel scan type iio:adc:ad7887 Fix channel reported endianness from cpu to big endian iio:accel:bma180 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:mag:mag3110 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5064 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5360 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5380 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5421 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5446 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5449 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5504 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5624r replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5686 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5755 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5764 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:ad5791 replacesIIO_STmacro with explicit entries to struct
scan_type
iio:dac:max517 drop specification of scan type as unused in this
driver.
iio:dac:mcp4725 drop specification of scan type as unused in this
driver.
iio:light:adjd_s311 replacesIIO_STmacro with explicit entries to
structscan_type
iio:light:tcs3472 replacesIIO_STmacro with explicit entries to
structscan_type
staging:iio:ad7606 replacesIIO_STmacro with explicit entries to
structscan_type
staging:iio:ad799x replacesIIO_STmacro with explicit entries to
structscan_type
iio:dropIIO_STmacrodrivers/iio/accel/bma180.c | 7 ++++++- drivers/iio/adc/ad7887.c | 16 ++++++++++++++-- drivers/iio/dac/ad5064.c | 7 ++++++- drivers/iio/dac/ad5360.c | 7 ++++++- drivers/iio/dac/ad5380.c | 7 ++++++- drivers/iio/dac/ad5421.c | 6 +++++- drivers/iio/dac/ad5446.c | 9 +++++++-- drivers/iio/dac/ad5449.c | 7 ++++++- drivers/iio/dac/ad5504.c | 6 +++++- drivers/iio/dac/ad5624r_spi.c | 7 ++++++- drivers/iio/dac/ad5686.c | 9 +++++++-- drivers/iio/dac/ad5755.c | 7 ++++++- drivers/iio/dac/ad5764.c | 7 ++++++- drivers/iio/dac/ad5791.c | 9 +++++++-- drivers/iio/dac/max517.c | 1 - drivers/iio/dac/mcp4725.c | 1 - drivers/iio/imu/adis16400_core.c | 7 ++++++-Quoted 9 lines
drivers/iio/light/adjd_s311.c | 7 ++++++- drivers/iio/light/tcs3472.c | 7 ++++++- drivers/iio/magnetometer/mag3110.c | 6 +++++- drivers/staging/iio/adc/ad7606_core.c | 7 ++++++- drivers/staging/iio/adc/ad799x_core.c | 8 +++++++- drivers/staging/iio/magnetometer/hmc5843.c | 7 ++++++- include/linux/iio/iio.h | 16 ++++++++++------ 24 files changed, 145 insertions(+), 33 deletions(-)--
To unsubscribe from this list: send the line "unsubscribe linux-iio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html