| Message ID | 1600753775-4745-1-git-send-email-cjhuang@codeaurora.org |
|---|---|
| State | Superseded |
| Headers | show |
| Series | [RFC,1/2] nl80211: add common API to configure SAR power limitations. | expand |
On Tue, 2020-09-22 at 13:49 +0800, Carl Huang wrote: > NL80211_CMD_SET_SAR_SPECS is added to configure SAR from > user space. NL80211_ATTR_SAR_SPEC is used to pass the SAR > power specification when used with NL80211_CMD_SET_SAR_SPECS. > > Wireless driver needs to register SAR type, supported frequency > ranges to wiphy, so user space can query it. The index in > frequency range is used to specify which sub band the power > limitation applies to. The SAR type is for compatibility, so later > other SAR mechanism can be implemented without breaking the user > space SAR applications. > > Normal process is user space quries the SAR capability, and typo: queries > gets the index of supported frequency ranges and associates the > power limitation with this index and sends to kernel. > > Here is an example of message send to kernel: > 8c 00 00 00 08 00 03 00 15 00 00 00 38 00 26 81 > 08 00 01 00 00 00 00 00 2c 00 02 80 14 00 01 80 > 05 00 02 00 00 00 00 00 05 00 01 00 38 00 00 00 > 14 00 02 80 05 00 02 00 01 00 00 00 05 00 01 00 > 48 00 00 00 Heh, that's not super useful, though that doesn't matter much I guess :) > NL80211_CMD_SET_SAR_SPECS: 0x8c > NL80211_ATTR_SAR_SPEC: 0x8126 (NLA_NESTED) > NL80211_SAR_ATTR_TYPE: 0x00 (NL80211_SAR_TYPE_POWER) > NL80211_SAR_ATTR_SPECS: 0x8002 (NLA_NESTED) > freq range 0 power: 0x38 in 0.25dbm unit (14dbm) > freq range 1 power: 0x48 in 0.25dbm unit (18dbm) that makes more sense :) > +/** > + * struct cfg80211_sar_sub_specs - sub specs limit > + * @power: value in 0.25dbm I guess the documentation should state that it's a power limit, or something? > + * @freq_range_index: index the power limitation applies to > + */ > +struct cfg80211_sar_sub_specs { > + u8 power; > + u8 freq_range_index; > +}; > + > +/** > + * struct cfg80211_sar_specs - sar limit specs > + * @type: it's set with power in 0.25dbm or other types > + * @num_sub_specs: number of sar sub specs > + * @sub_specs: memory to hold the sar sub specs > + */ > +struct cfg80211_sar_specs { > + enum nl80211_sar_type type; > + u16 num_sub_specs; > + struct cfg80211_sar_sub_specs *sub_specs; Wouldn't it make more sense to just have sub_specs[] and inline? The pointers here are bigger than many of the sub-specs I guess... Also, is num_sub_specs even needed? It seems we could just require that userspace sends all of them all the time? > +struct cfg80211_sar_freq_ranges { > + u8 index; Does an index here make sense? > + u32 start_freq; > + u32 end_freq; > +}; > + > +/** > + * struct cfg80211_sar_capa - sar limit capability > + * @type: it's set via power in 0.25dbm or other types > + * @num_freq_ranges: number of frequency ranges > + * @chan_ranges: memory to hold the channel ranges. > + */ > +struct cfg80211_sar_capa { > + enum nl80211_sar_type type; > + u8 num_freq_ranges; > + const struct cfg80211_sar_freq_ranges *freq_ranges; You end up with an array here, so some kind of index is implicit, the array index. I don't see why you'd want to have a separate index *within* the frequency range, then you just have to check those make sense, etc.? > + * @NL80211_ATTR_SAR_SPEC: SAR power limitation specification when > + * used with %NL80211_CMD_SET_SAR_SPECS. It contains array of > + * %nl80211_sar_specs_attrs. I think I know what you mean (a nested array where each item has attributes from &enum nl80211_sar_specs_attrs), but could state that a bit clearer. > +#define NL80211_SAR_ALL_FREQ_RNAGES 0xff typo > +#define NUM_MAX_NL80211_SAR_FREQ_RANGES 0xfe but I'm not sure what these are used for in the first place, they seem more like internal implementation details? > +/** > + * nl80211_sar_specs_attrs - Attributes for SAR power limit specs > + * > + * @NL80211_SAR_ATTR_SPECS_POWER: Required (u32)value to specify the actual > + * power limit value in units of 0.25 dBm if type is > + * NL80211_SAR_TYPE_POWER. (i.e., a value of 44 represents 11 dBm) > + * > + * @NL80211_SAR_ATTR_SPECS_FREQ_RANGES_INDEX: optional (u32) value to specify the > + * index of exported freq ranges table. If this attribute is not present, then > + * the power is applied to all freq ranges, i.e, all bands Especially since this says "[i]f this attribute is not present". Does that make sense though? It seems possible to just use the nesting entry as the index, and just require userspace to set up all indexes if it wants to apply to all of them, why bother having to deal with that case in the kernel and perhaps even drivers? > + * @NL80211_SAR_ATTR_SPECS_START_FREQ: Required (u32) value to specify the start > + * frequency of this range to register SAR capability to wihpy and the unit typo: wiphy > + * is kHZ kHz > + * > + * @NL80211_SAR_ATTR_SPECS_END_FREQ: Required (u32) value to specify the end frequency > + * of this range to register SAR capability to wiphy and the unit is kHZ kHz But hm, are you mixing the capability advertisement and the actual request into the same attributes? Perhaps that makes sense, but the documentation could be a bit clearer IMHO. > +++ b/net/mac80211/cfg.c The mac80211 bits could probably be in a separate patch, just for clarity. > +static const struct nla_policy > +sar_specs_policy[NL80211_SAR_ATTR_SPECS_MAX + 1] = { > + [NL80211_SAR_ATTR_SPECS_POWER] = { .type = NLA_U8 }, double indentation? > + [NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX] = > + NLA_POLICY_MAX(NLA_U8, NUM_MAX_NL80211_SAR_FREQ_RANGES), I don't see a good reason to limit to 0xfe? The kernel could just use a wider datatype internally. Anyway, I'm not sure the attribute is even needed, see above. > @@ -1856,6 +1870,8 @@ static int nl80211_add_commands_unsplit(struct cfg80211_registered_device *rdev, > goto nla_put_failure; > } > > + CMD(set_sar_specs, SET_SAR_SPECS); Don't add to _unsplit() please. > + specs = nla_nest_start_noflag(msg, NL80211_SAR_ATTR_SPECS); Don't use _noflag() please, that's just for backward compatibility. > + if (!specs) > + goto fail; > + > + /* report supported freq_ranges */ > + for (i = 0; i < num_freq_ranges; i++) { > + sub_freq_range = nla_nest_start_noflag(msg, i + 1); Same here. > @@ -2598,6 +2663,9 @@ static int nl80211_send_wiphy(struct cfg80211_registered_device *rdev, > if (nl80211_put_tid_config_support(rdev, msg)) > goto nla_put_failure; > > + if (nl80211_put_sar_specs(rdev, msg)) > + goto nla_put_failure; Perhaps better to have a separate message (case entry) since it might get fairly big? > + sar_spec = kzalloc(sizeof(*sar_spec) + > + specs * sizeof(struct cfg80211_sar_sub_specs), > + GFP_KERNEL); > + if (!sar_spec) > + return -ENOMEM; > + > + sar_spec->sub_specs = (struct cfg80211_sar_sub_specs *) > + ((char *)sar_spec + sizeof(*sar_spec)); If you anyway set the pointer there then you can actually just make it a [] array instead of having the pointer in the first place ... would be equivalent but save you from these contortions :) > + specs = 0; > + sar_spec->type = type; > + > + nla_for_each_nested(spec_list, tb[NL80211_SAR_ATTR_SPECS], rem) { > + if (nla_parse(spec, > + NL80211_SAR_ATTR_SPECS_MAX, > + nla_data(spec_list), > + nla_len(spec_list), > + sar_specs_policy, > + NULL)) { > + err = -EINVAL; > + goto error; > + } > + > + /* for power type, power value must be presented */ > + if (!spec[NL80211_SAR_ATTR_SPECS_POWER] && > + type == NL80211_SAR_TYPE_POWER) { > + err = -EINVAL; > + goto error; > + } Somewhere you should probably also check that the type matches the advertised type. > + power = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_POWER]); > + sar_spec->sub_specs[specs].power = power; > + > + /* if NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX isn't present, > + * then the power applies to all bands. But it's only valid > + * for the first entry. > + */ > + if (!spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]) { > + if (specs) { > + err = -EINVAL; > + goto error; > + } else { > + sar_spec->sub_specs[specs].freq_range_index = > + NL80211_SAR_ALL_FREQ_RNAGES; > + specs++; > + break; > + } > + } Yeah... not sure I see this as being worth it, vs. userspace just filling all the possible entries. > + index = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]); > + sar_spec->sub_specs[specs].freq_range_index = index; > + specs++; > + } > + > + sar_spec->num_sub_specs = specs; > + > + rdev->cur_cmd_info = info; > + if (rdev->ops->set_sar_specs) > + err = rdev_set_sar_specs(rdev, sar_spec); > + else > + err = -EOPNOTSUPP; Might be worth checking this early and avoiding the allocation and all? OTOH, that gives you error checking even if the driver was completely confused and advertised sar specs without the method ... johannes
From: kuabhs@chromium.org These patches do not apply cleanly on kernel 5.4. I fixed the merged conflicts and verified. I verified these APIs using power_manager in chromeos, the APIs works fine. Tested-by: Abhishek Kumar <kuabhs@chromium.org> -Abhishek
From: kuabhs@chromium.org There are few more additional comments here. > + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is > + * passed using %NL80211_ATTR_SAR_SPEC. > + * This command requires NL80211_ATTR_IFINDEX, probably should be better to add this in the comment ? > +static int > +nl80211_put_sar_specs(struct cfg80211_registered_device *rdev, > + struct sk_buff *msg) > +{ > + struct nlattr *sar_capa, *specs, *sub_freq_range; > + u8 num_freq_ranges; > + int i; > + > + if (!rdev->wiphy.sar_capa) > + return 0; > + > + num_freq_ranges = rdev->wiphy.sar_capa->num_freq_ranges; > + > + sar_capa = nla_nest_start(msg, NL80211_ATTR_SAR_SPEC); > + if (!sar_capa) > + return -ENOSPC; I see some places nla_nest_start_noflag being used and here nla_nest_start. Whats the specific reason to do that ? In the newer Kernel versions, I believe nla_nest_start is preferred. > + power = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_POWER]); > + sar_spec->sub_specs[specs].power = power; > + > + /* if NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX isn't present, > + * then the power applies to all bands. But it's only valid > + * for the first entry. > + */ > + if (!spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]) { > + if (specs) { > + err = -EINVAL; > + goto error; > + } else { > + sar_spec->sub_specs[specs].freq_range_index = > + NL80211_SAR_ALL_FREQ_RNAGES; > + specs++; > + break; > + } > + } Here I see you are assigning same power to all freq band if for the first band the freq index is not provided. Is there any specific reason to only check the first here ? Probably this logic should move into specific drivers. Thoughts ? -Abhishek
On 2020-10-31 10:46, Abhishek Kumar wrote: > From: kuabhs@chromium.org > > There are few more additional comments here. >> + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is >> + * passed using %NL80211_ATTR_SAR_SPEC. >> + * > > This command requires NL80211_ATTR_IFINDEX, probably should be better > to add > this in the comment ? > Per Johannes's comments, this explicit index is not required. Are you fine with it? Instead, user-space application records the array index when querying the SAR capability. When set, the nested array index will be used to set the power. This requires user-space has a strict mapping of index. and NL80211_ATTR_IFINDEX is to be removed. >> +static int >> +nl80211_put_sar_specs(struct cfg80211_registered_device *rdev, >> + struct sk_buff *msg) >> +{ >> + struct nlattr *sar_capa, *specs, *sub_freq_range; >> + u8 num_freq_ranges; >> + int i; >> + >> + if (!rdev->wiphy.sar_capa) >> + return 0; >> + >> + num_freq_ranges = rdev->wiphy.sar_capa->num_freq_ranges; >> + >> + sar_capa = nla_nest_start(msg, NL80211_ATTR_SAR_SPEC); >> + if (!sar_capa) >> + return -ENOSPC; > > I see some places nla_nest_start_noflag being used and here > nla_nest_start. > Whats the specific reason to do that ? In the newer Kernel versions, I > believe > nla_nest_start is preferred. > This will be addressed in next version. >> + power = >> nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_POWER]); >> + sar_spec->sub_specs[specs].power = power; >> + >> + /* if NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX isn't >> present, >> + * then the power applies to all bands. But it's only >> valid >> + * for the first entry. >> + */ >> + if (!spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]) { >> + if (specs) { >> + err = -EINVAL; >> + goto error; >> + } else { >> + >> sar_spec->sub_specs[specs].freq_range_index = >> + NL80211_SAR_ALL_FREQ_RNAGES; >> + specs++; >> + break; >> + } >> + } > > Here I see you are assigning same power to all freq band if for the > first band > the freq index is not provided. Is there any specific reason to only > check the first > here ? Probably this logic should move into specific drivers. Thoughts > ? > This logic will be removed per Johannes's comments. Please read Johannes's comments. If you agree with him and has no other advices, then I will post the second version of it. > -Abhishek > > > _______________________________________________ > ath10k mailing list > ath10k@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/ath10k
On Tue, 2020-11-03 at 10:34 +0800, Carl Huang wrote: > On 2020-10-31 10:46, Abhishek Kumar wrote: > > From: kuabhs@chromium.org > > > > There are few more additional comments here. > > > + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is > > > + * passed using %NL80211_ATTR_SAR_SPEC. > > > + * > > > > This command requires NL80211_ATTR_IFINDEX, probably should be better > > to add > > this in the comment ? > > > Per Johannes's comments, this explicit index is not required. Are you > fine with it? > > Instead, user-space application records the array index when querying > the SAR > capability. When set, the nested array index will be used to set the > power. > This requires user-space has a strict mapping of index. and > NL80211_ATTR_IFINDEX > is to be removed. Wait, what? The IFINDEX is still required, that identifies the interface - though it probably shouldn't be required, this should be per wiphy, so you could also use ATTR_WIPHY_IDX? You're thinking of ... something else. NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX perhaps? johannes
On Tue, Nov 3, 2020 at 5:15 AM Johannes Berg <johannes@sipsolutions.net> wrote: > > On Tue, 2020-11-03 at 10:34 +0800, Carl Huang wrote: > > On 2020-10-31 10:46, Abhishek Kumar wrote: > > > From: kuabhs@chromium.org > > > > > > There are few more additional comments here. > > > > + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is > > > > + * passed using %NL80211_ATTR_SAR_SPEC. > > > > + * > > > > > > This command requires NL80211_ATTR_IFINDEX, probably should be better > > > to add > > > this in the comment ? > > > > > Per Johannes's comments, this explicit index is not required. Are you > > fine with it? > > > > Instead, user-space application records the array index when querying > > the SAR > > capability. When set, the nested array index will be used to set the > > power. > > This requires user-space has a strict mapping of index. and > > NL80211_ATTR_IFINDEX > > is to be removed. > > Wait, what? The IFINDEX is still required, that identifies the interface > - though it probably shouldn't be required, this should be per wiphy, so > you could also use ATTR_WIPHY_IDX? > > You're thinking of ... something > else. NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX perhaps? Yes, probably the index mapping that you are talking about is using NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX . I think Johannes has already covered most of the comments. I have one last one mentioned below. > +nl80211_put_sar_specs(struct cfg80211_registered_device *rdev, > + struct sk_buff *msg) > +{ > + struct nlattr *sar_capa, *specs, *sub_freq_range; > + u8 num_freq_ranges; > + int i; > + > + if (!rdev->wiphy.sar_capa) > + return 0; > + > + num_freq_ranges = rdev->wiphy.sar_capa->num_freq_ranges; > + > + sar_capa = nla_nest_start(msg, NL80211_ATTR_SAR_SPEC); > + if (!sar_capa) > + return -ENOSPC; > + > + if (nla_put_u16(msg, NL80211_SAR_ATTR_TYPE, rdev->wiphy.sar_capa->type)) > + goto fail; The nla_put_u16 NL80211_SAR_ATTR_TYPE here uses u16 whereas in the to get this property below it uses nla_get_u32 . I think this should be consistent ? > + if (!tb[NL80211_SAR_ATTR_TYPE]) > + return -EINVAL; > + > + type = nla_get_u32(tb[NL80211_SAR_ATTR_TYPE]); > + > + if (!tb[NL80211_SAR_ATTR_SPECS]) > + return -EINVAL; As mentioned above the get and put for NL80211_SAR_ATTR_TYPE is not consistent. -Abhishek
On 2020-11-04 09:17, Abhishek Kumar wrote: > On Tue, Nov 3, 2020 at 5:15 AM Johannes Berg > <johannes@sipsolutions.net> wrote: >> >> On Tue, 2020-11-03 at 10:34 +0800, Carl Huang wrote: >> > On 2020-10-31 10:46, Abhishek Kumar wrote: >> > > From: kuabhs@chromium.org >> > > >> > > There are few more additional comments here. >> > > > + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is >> > > > + * passed using %NL80211_ATTR_SAR_SPEC. >> > > > + * >> > > >> > > This command requires NL80211_ATTR_IFINDEX, probably should be better >> > > to add >> > > this in the comment ? >> > > >> > Per Johannes's comments, this explicit index is not required. Are you >> > fine with it? >> > >> > Instead, user-space application records the array index when querying >> > the SAR >> > capability. When set, the nested array index will be used to set the >> > power. >> > This requires user-space has a strict mapping of index. and >> > NL80211_ATTR_IFINDEX >> > is to be removed. >> >> Wait, what? The IFINDEX is still required, that identifies the >> interface >> - though it probably shouldn't be required, this should be per wiphy, >> so >> you could also use ATTR_WIPHY_IDX? >> >> You're thinking of ... something >> else. NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX perhaps? > > Yes, probably the index mapping that you are talking about is using > NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX . > You're right. I just remembered Johannes's comments to remove the NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX and made mistake here. > I think Johannes has already covered most of the comments. I have one > last one mentioned below. > >> +nl80211_put_sar_specs(struct cfg80211_registered_device *rdev, >> + struct sk_buff *msg) >> +{ >> + struct nlattr *sar_capa, *specs, *sub_freq_range; >> + u8 num_freq_ranges; >> + int i; >> + >> + if (!rdev->wiphy.sar_capa) >> + return 0; >> + >> + num_freq_ranges = rdev->wiphy.sar_capa->num_freq_ranges; >> + >> + sar_capa = nla_nest_start(msg, NL80211_ATTR_SAR_SPEC); >> + if (!sar_capa) >> + return -ENOSPC; >> + >> + if (nla_put_u16(msg, NL80211_SAR_ATTR_TYPE, >> rdev->wiphy.sar_capa->type)) >> + goto fail; > > The nla_put_u16 NL80211_SAR_ATTR_TYPE here uses u16 whereas in the to > get this property below it uses nla_get_u32 . I think this should be > consistent ? > >> + if (!tb[NL80211_SAR_ATTR_TYPE]) >> + return -EINVAL; >> + >> + type = nla_get_u32(tb[NL80211_SAR_ATTR_TYPE]); >> + >> + if (!tb[NL80211_SAR_ATTR_SPECS]) >> + return -EINVAL; > > As mentioned above the get and put for NL80211_SAR_ATTR_TYPE is not > consistent. > Correct. Thanks for pointing out and it will be fixed in next version. > -Abhishek > > _______________________________________________ > ath10k mailing list > ath10k@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/ath10k
On 2020-09-28 20:36, Johannes Berg wrote: > On Tue, 2020-09-22 at 13:49 +0800, Carl Huang wrote: >> NL80211_CMD_SET_SAR_SPECS is added to configure SAR from >> user space. NL80211_ATTR_SAR_SPEC is used to pass the SAR >> power specification when used with NL80211_CMD_SET_SAR_SPECS. >> >> Wireless driver needs to register SAR type, supported frequency >> ranges to wiphy, so user space can query it. The index in >> frequency range is used to specify which sub band the power >> limitation applies to. The SAR type is for compatibility, so later >> other SAR mechanism can be implemented without breaking the user >> space SAR applications. >> >> Normal process is user space quries the SAR capability, and > > typo: queries > sure, I'll fix it in v2. >> gets the index of supported frequency ranges and associates the >> power limitation with this index and sends to kernel. >> >> Here is an example of message send to kernel: >> 8c 00 00 00 08 00 03 00 15 00 00 00 38 00 26 81 >> 08 00 01 00 00 00 00 00 2c 00 02 80 14 00 01 80 >> 05 00 02 00 00 00 00 00 05 00 01 00 38 00 00 00 >> 14 00 02 80 05 00 02 00 01 00 00 00 05 00 01 00 >> 48 00 00 00 > > Heh, that's not super useful, though that doesn't matter much I guess > :) > >> NL80211_CMD_SET_SAR_SPECS: 0x8c >> NL80211_ATTR_SAR_SPEC: 0x8126 (NLA_NESTED) >> NL80211_SAR_ATTR_TYPE: 0x00 (NL80211_SAR_TYPE_POWER) >> NL80211_SAR_ATTR_SPECS: 0x8002 (NLA_NESTED) >> freq range 0 power: 0x38 in 0.25dbm unit (14dbm) >> freq range 1 power: 0x48 in 0.25dbm unit (18dbm) > > that makes more sense :) > As this is an explanation of the message above, is it good to keep the message? >> +/** >> + * struct cfg80211_sar_sub_specs - sub specs limit >> + * @power: value in 0.25dbm > > I guess the documentation should state that it's a power limit, or > something? > >> + * @freq_range_index: index the power limitation applies to >> + */ >> +struct cfg80211_sar_sub_specs { >> + u8 power; >> + u8 freq_range_index; >> +}; >> + >> +/** >> + * struct cfg80211_sar_specs - sar limit specs >> + * @type: it's set with power in 0.25dbm or other types >> + * @num_sub_specs: number of sar sub specs >> + * @sub_specs: memory to hold the sar sub specs >> + */ >> +struct cfg80211_sar_specs { >> + enum nl80211_sar_type type; >> + u16 num_sub_specs; >> + struct cfg80211_sar_sub_specs *sub_specs; > > Wouldn't it make more sense to just have sub_specs[] and inline? The > pointers here are bigger than many of the sub-specs I guess... > > Also, is num_sub_specs even needed? It seems we could just require that > userspace sends all of them all the time? > Agree as long as user space sends all of them. >> +struct cfg80211_sar_freq_ranges { >> + u8 index; > > Does an index here make sense? > With agreement from Google, it's OK to remove it. >> + u32 start_freq; >> + u32 end_freq; >> +}; >> + >> +/** >> + * struct cfg80211_sar_capa - sar limit capability >> + * @type: it's set via power in 0.25dbm or other types >> + * @num_freq_ranges: number of frequency ranges >> + * @chan_ranges: memory to hold the channel ranges. >> + */ >> +struct cfg80211_sar_capa { >> + enum nl80211_sar_type type; >> + u8 num_freq_ranges; >> + const struct cfg80211_sar_freq_ranges *freq_ranges; > > You end up with an array here, so some kind of index is implicit, the > array index. I don't see why you'd want to have a separate index > *within* the frequency range, then you just have to check those make > sense, etc.? > >> + * @NL80211_ATTR_SAR_SPEC: SAR power limitation specification when >> + * used with %NL80211_CMD_SET_SAR_SPECS. It contains array of >> + * %nl80211_sar_specs_attrs. > > I think I know what you mean (a nested array where each item has > attributes from &enum nl80211_sar_specs_attrs), but could state that a > bit clearer. > I'll try to describe it more clear in v2. >> +#define NL80211_SAR_ALL_FREQ_RNAGES 0xff > > typo > sure, will fix in v2. >> +#define NUM_MAX_NL80211_SAR_FREQ_RANGES 0xfe > > but I'm not sure what these are used for in the first place, they seem > more like internal implementation details? > In v1 design, userspace doesn't need to specify all the sub bands if the a power limitation applies to all sub bands, so NL80211_SAR_ALL_FREQ_RNAGES is defined for this purpose. To distinguish from NL80211_SAR_ALL_FREQ_RNAGES, the maximum number of freq ranges should not exceed (NL80211_SAR_ALL_FREQ_RNAGES - 1), so define NUM_MAX_NL80211_SAR_FREQ_RANGES as 0xfe. It's right these two definitions are internal, userspace doesn't use them. Again as user-space will sends all the ranges, let me remove these 2 definitions. >> +/** >> + * nl80211_sar_specs_attrs - Attributes for SAR power limit specs >> + * >> + * @NL80211_SAR_ATTR_SPECS_POWER: Required (u32)value to specify the >> actual >> + * power limit value in units of 0.25 dBm if type is >> + * NL80211_SAR_TYPE_POWER. (i.e., a value of 44 represents 11 dBm) >> + * >> + * @NL80211_SAR_ATTR_SPECS_FREQ_RANGES_INDEX: optional (u32) value to >> specify the >> + * index of exported freq ranges table. If this attribute is not >> present, then >> + * the power is applied to all freq ranges, i.e, all bands > > Especially since this says "[i]f this attribute is not present". > > Does that make sense though? It seems possible to just use the nesting > entry as the index, and just require userspace to set up all indexes if > it wants to apply to all of them, why bother having to deal with that > case in the kernel and perhaps even drivers? > Ok, let me drop it in v2. Johannes, want to confirm with you "nesting entry as the index". Do you mean that kernel take the first entry as index 0, then second entry as index 1 and etc? >> + * @NL80211_SAR_ATTR_SPECS_START_FREQ: Required (u32) value to >> specify the start >> + * frequency of this range to register SAR capability to wihpy and >> the unit > > typo: wiphy > >> + * is kHZ > > kHz > Will fix it in v2. >> + * >> + * @NL80211_SAR_ATTR_SPECS_END_FREQ: Required (u32) value to specify >> the end frequency >> + * of this range to register SAR capability to wiphy and the unit is >> kHZ > > kHz > > But hm, are you mixing the capability advertisement and the actual > request into the same attributes? Perhaps that makes sense, but the > documentation could be a bit clearer IMHO. > >> +++ b/net/mac80211/cfg.c > > The mac80211 bits could probably be in a separate patch, just for > clarity. > >> +static const struct nla_policy >> +sar_specs_policy[NL80211_SAR_ATTR_SPECS_MAX + 1] = { >> + [NL80211_SAR_ATTR_SPECS_POWER] = { .type = NLA_U8 }, > > double indentation? > >> + [NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX] = >> + NLA_POLICY_MAX(NLA_U8, NUM_MAX_NL80211_SAR_FREQ_RANGES), > > I don't see a good reason to limit to 0xfe? The kernel could just use a > wider datatype internally. > > Anyway, I'm not sure the attribute is even needed, see above. > If userspace has to sends all the sub bands, then NL80211_SAR_ALL_FREQ_RNAGES is not required, and therefore NUM_MAX_NL80211_SAR_FREQ_RANGES is also not required. I'll delete these 2 definitions, and change policy to { .type = NLA_U8 } >> @@ -1856,6 +1870,8 @@ static int nl80211_add_commands_unsplit(struct >> cfg80211_registered_device *rdev, >> goto nla_put_failure; >> } >> >> + CMD(set_sar_specs, SET_SAR_SPECS); > > Don't add to _unsplit() please. > Do you mean add it to below block? if (state->split) { ... CMD(set_sar_specs, SET_SAR_SPECS); } >> + specs = nla_nest_start_noflag(msg, NL80211_SAR_ATTR_SPECS); > > Don't use _noflag() please, that's just for backward compatibility. > sure. >> + if (!specs) >> + goto fail; >> + >> + /* report supported freq_ranges */ >> + for (i = 0; i < num_freq_ranges; i++) { >> + sub_freq_range = nla_nest_start_noflag(msg, i + 1); > > Same here. > >> @@ -2598,6 +2663,9 @@ static int nl80211_send_wiphy(struct >> cfg80211_registered_device *rdev, >> if (nl80211_put_tid_config_support(rdev, msg)) >> goto nla_put_failure; >> >> + if (nl80211_put_sar_specs(rdev, msg)) >> + goto nla_put_failure; > > Perhaps better to have a separate message (case entry) since it might > get fairly big? > >> + sar_spec = kzalloc(sizeof(*sar_spec) + >> + specs * sizeof(struct cfg80211_sar_sub_specs), >> + GFP_KERNEL); >> + if (!sar_spec) >> + return -ENOMEM; >> + >> + sar_spec->sub_specs = (struct cfg80211_sar_sub_specs *) >> + ((char *)sar_spec + sizeof(*sar_spec)); > > If you anyway set the pointer there then you can actually just make it > a > [] array instead of having the pointer in the first place ... would be > equivalent but save you from these contortions :) > Right. >> + specs = 0; >> + sar_spec->type = type; >> + >> + nla_for_each_nested(spec_list, tb[NL80211_SAR_ATTR_SPECS], rem) { >> + if (nla_parse(spec, >> + NL80211_SAR_ATTR_SPECS_MAX, >> + nla_data(spec_list), >> + nla_len(spec_list), >> + sar_specs_policy, >> + NULL)) { >> + err = -EINVAL; >> + goto error; >> + } >> + >> + /* for power type, power value must be presented */ >> + if (!spec[NL80211_SAR_ATTR_SPECS_POWER] && >> + type == NL80211_SAR_TYPE_POWER) { >> + err = -EINVAL; >> + goto error; >> + } > > Somewhere you should probably also check that the type matches the > advertised type. > Agree. >> + power = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_POWER]); >> + sar_spec->sub_specs[specs].power = power; >> + >> + /* if NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX isn't present, >> + * then the power applies to all bands. But it's only valid >> + * for the first entry. >> + */ >> + if (!spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]) { >> + if (specs) { >> + err = -EINVAL; >> + goto error; >> + } else { >> + sar_spec->sub_specs[specs].freq_range_index = >> + NL80211_SAR_ALL_FREQ_RNAGES; >> + specs++; >> + break; >> + } >> + } > > Yeah... not sure I see this as being worth it, vs. userspace just > filling all the possible entries. > Yes, let me remove this logic in v2. >> + index = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]); >> + sar_spec->sub_specs[specs].freq_range_index = index; >> + specs++; >> + } >> + >> + sar_spec->num_sub_specs = specs; >> + >> + rdev->cur_cmd_info = info; >> + if (rdev->ops->set_sar_specs) >> + err = rdev_set_sar_specs(rdev, sar_spec); >> + else >> + err = -EOPNOTSUPP; > > Might be worth checking this early and avoiding the allocation and all? > > OTOH, that gives you error checking even if the driver was completely > confused and advertised sar specs without the method ... > Here, it checks whether mac80211 layer implements the function. It's ieee80211_set_sar_specs to check whether driver has implemented set_sar function. I'll move the check to the beginning of this function. > johannes
On Wed, Nov 4, 2020 at 12:44 AM Carl Huang <cjhuang@codeaurora.org> wrote: > On 2020-09-28 20:36, Johannes Berg wrote: > > On Tue, 2020-09-22 at 13:49 +0800, Carl Huang wrote: > >> +struct cfg80211_sar_freq_ranges { > >> + u8 index; > > > > Does an index here make sense? > > > With agreement from Google, it's OK to remove it. I'm not sure "Google" is the arbiter of the nl80211 API, even if we are the current planned users ;) But I think I agree with Johannes, that given the other plans (user space must send all bands all the time; dropping the "apply to all bands" support), an index isn't really necessary in either the user space API or the internal representation handed down to drivers. All bands should be specified, in order. Brian > >> + u32 start_freq; > >> + u32 end_freq; > >> +};
Hi, On Tue, Sep 22, 2020 at 01:49:35PM +0800, Carl Huang wrote: > ath10k assigns ath10k_mac_set_sar_specs to ath10k_ops, and > this function is called when user space application calls > NL80211_CMD_SET_SAR_SPECS. ath10k also registers SAR type, > and supported frequency ranges to wiphy so user space can > query SAR capabilities. > > ath10k_mac_set_sar_specs further sets the power to firmware > to limit the TX power. > > This feature is controlled by hw parameter: dynamic_sar_support. > > Signed-off-by: Carl Huang <cjhuang@codeaurora.org> > --- > diff --git a/drivers/net/wireless/ath/ath10k/mac.c b/drivers/net/wireless/ath/ath10k/mac.c > index 2e3eb5b..830c61f 100644 > --- a/drivers/net/wireless/ath/ath10k/mac.c > +++ b/drivers/net/wireless/ath/ath10k/mac.c > @@ -81,6 +81,17 @@ static struct ieee80211_rate ath10k_rates_rev2[] = { > { .bitrate = 540, .hw_value = ATH10K_HW_RATE_OFDM_54M }, > }; > > +static const struct cfg80211_sar_freq_ranges ath10k_sar_freq_ranges[] = { > + { .index = 0, .start_freq = 2412000, .end_freq = 2484000 }, 2412 MHz is a center frequency, but other parts of the nl80211 API use band edges. For example: * @NL80211_ATTR_FREQ_RANGE_START: starting frequencry for the regulatory * rule in KHz. This is not a center of frequency but an actual regulatory * band edge. * @NL80211_ATTR_FREQ_RANGE_END: ending frequency for the regulatory rule * in KHz. This is not a center a frequency but an actual regulatory * band edge. Seems like we should improve the nl80211.h docs (patch 1) and make these edges (this patch). > + { .index = 1, .start_freq = 2484000, .end_freq = 5865000 }, > +}; > + > +static const struct cfg80211_sar_capa ath10k_sar_capa = { > + .type = NL80211_SAR_TYPE_POWER, > + .num_freq_ranges = (ARRAY_SIZE(ath10k_sar_freq_ranges)), > + .freq_ranges = &ath10k_sar_freq_ranges[0], > +}; > + > #define ATH10K_MAC_FIRST_OFDM_RATE_IDX 4 > > #define ath10k_a_rates (ath10k_rates + ATH10K_MAC_FIRST_OFDM_RATE_IDX) > @@ -2880,6 +2891,95 @@ static int ath10k_mac_vif_recalc_txbf(struct ath10k *ar, > return 0; > } > > +static bool ath10k_mac_is_connected(struct ath10k *ar) > +{ > + struct ath10k_vif *arvif; > + > + list_for_each_entry(arvif, &ar->arvifs, list) { > + if (arvif->is_up && arvif->vdev_type == WMI_VDEV_TYPE_STA) > + return true; > + } > + > + return false; > +} > + > +int ath10k_mac_set_sar_power(struct ath10k *ar) This function should be static. > +{ > + int ret; > + > + if (!ar->hw_params.dynamic_sar_support) > + return 0; return -EOPNOTSUPP ? > + > + if (ar->tx_power_2g_limit == 0 || ar->tx_power_5g_limit == 0) ath10k_mac_txpower_recalc() doesn't care about this -- why should you? This also seems especially weird, because one of the 2 could be valid nonzero values, and yet you're silently rejecting it. Regardless, the following seems wrong: > + return 0; This should probably be an error. > + > + if (!ath10k_mac_is_connected(ar)) > + return 0; Note to self (since this wasn't obvious to me the first read-through): you're calling this function from ath10k_bss_assoc() too, so even if you weren't connected the first time around, it'll get called later. > + > + ret = ath10k_wmi_pdev_set_param(ar, > + ar->wmi.pdev_param->txpower_limit2g, > + ar->tx_power_2g_limit); > + if (ret) { > + ath10k_warn(ar, "failed to set 2.4G txpower %d: %d\n", > + ar->tx_power_2g_limit, ret); > + return ret; > + } > + > + ret = ath10k_wmi_pdev_set_param(ar, > + ar->wmi.pdev_param->txpower_limit5g, > + ar->tx_power_5g_limit); > + if (ret) { > + ath10k_warn(ar, "failed to set 5G txpower %d: %d\n", > + ar->tx_power_5g_limit, ret); > + return ret; > + } Hmm, so these are the same params configured by ath10k_mac_txpower_recalc(), except that we're not taking into account the limitations in ath10k_mac_txpower_recalc() (and vice versa) -- isn't that bad? Should we be merging the SAR limitation into ath10k_mac_txpower_recalc() and calling that instead? Brian > + > + ath10k_dbg(ar, ATH10K_DBG_MAC, "set txpower 2G:%d, 5G:%d successfully\n", > + ar->tx_power_2g_limit, ar->tx_power_5g_limit); > + > + return ret; > +} > +
On Mon, Sep 28, 2020 at 02:36:09PM +0200, Johannes Berg wrote: > > +#define NUM_MAX_NL80211_SAR_FREQ_RANGES 0xfe > > but I'm not sure what these are used for in the first place, they seem > more like internal implementation details? I think the MAX value does have some utility in the API -- as mentioned in other comments, if we're requiring that user space must SET all ranges at the same time, then we need an expected way for user space to SET a "don't care" or "MAX" or "null" value for a band. So if there's some new band (e.g., 6 GHz?) that user space was not previously aware of, it will know to use this placeholder. I think that's kind of approximately what the purpose of this was? It's not super clear in the documentation, so maybe that should be clarified in writing in v2. Brian
Hi Carl, Sorry, I lied; I have a few more notes after spending another day looking at this: On Tue, Sep 22, 2020 at 01:49:34PM +0800, Carl Huang wrote: > --- a/include/net/cfg80211.h > +++ b/include/net/cfg80211.h > @@ -1663,6 +1663,55 @@ struct station_info { > +/** > + * @struct cfg80211_sar_chan_ranges - sar frequency ranges > + * @index: the index of this range. It's used to specify > + * the frequency range when setting SAR power limitation > + * @start_freq: start channel frequency in kHZ. For example, > + * 2.4G channel 1 is represented as 2412000 > + * @end_freq: end channel frequency in kHZ If you accept my comments in nl80211.h, you'll want to change this too. > + */ > +struct cfg80211_sar_freq_ranges { > + u8 index; > + u32 start_freq; > + u32 end_freq; > +}; > --- a/include/uapi/linux/nl80211.h > +++ b/include/uapi/linux/nl80211.h > + * @NL80211_SAR_ATTR_SPECS_START_FREQ: Required (u32) value to specify the start > + * frequency of this range to register SAR capability to wihpy and the unit > + * is kHZ > + * > + * @NL80211_SAR_ATTR_SPECS_END_FREQ: Required (u32) value to specify the end frequency > + * of this range to register SAR capability to wiphy and the unit is kHZ The documentation here isn't clear whether these are center frequencies or band edges. The cfg80211 comments do though (center freq). However, this is inconsistent with NL80211_ATTR_FREQ_RANGE_START and NL80211_ATTR_FREQ_RANGE_END -- I'd suggest being consistent? > --- a/net/mac80211/cfg.c > +++ b/net/mac80211/cfg.c > @@ -15331,6 +15496,14 @@ static const struct genl_ops nl80211_ops[] = { > .internal_flags = NL80211_FLAG_NEED_NETDEV | > NL80211_FLAG_NEED_RTNL, > }, > + { > + .cmd = NL80211_CMD_SET_SAR_SPECS, > + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, > + .doit = nl80211_set_sar_specs, > + .flags = GENL_UNS_ADMIN_PERM, > + .internal_flags = NL80211_FLAG_NEED_NETDEV_UP | Is there a reason this needs to be UP? CMD_SET_WIPHY (which can also configure TX power, a little differently) does not. Seems like this could just be NL80211_FLAG_NEED_NETDEV -- or maybe not even that, if we switch this to a WIPHY command like Johannes noted. Brian > + NL80211_FLAG_NEED_RTNL, > + }, > };
On 2020-11-05 07:11, Brian Norris wrote: > Hi, > > On Tue, Sep 22, 2020 at 01:49:35PM +0800, Carl Huang wrote: >> ath10k assigns ath10k_mac_set_sar_specs to ath10k_ops, and >> this function is called when user space application calls >> NL80211_CMD_SET_SAR_SPECS. ath10k also registers SAR type, >> and supported frequency ranges to wiphy so user space can >> query SAR capabilities. >> >> ath10k_mac_set_sar_specs further sets the power to firmware >> to limit the TX power. >> >> This feature is controlled by hw parameter: dynamic_sar_support. >> >> Signed-off-by: Carl Huang <cjhuang@codeaurora.org> >> --- > >> diff --git a/drivers/net/wireless/ath/ath10k/mac.c >> b/drivers/net/wireless/ath/ath10k/mac.c >> index 2e3eb5b..830c61f 100644 >> --- a/drivers/net/wireless/ath/ath10k/mac.c >> +++ b/drivers/net/wireless/ath/ath10k/mac.c >> @@ -81,6 +81,17 @@ static struct ieee80211_rate ath10k_rates_rev2[] = >> { >> { .bitrate = 540, .hw_value = ATH10K_HW_RATE_OFDM_54M }, >> }; >> >> +static const struct cfg80211_sar_freq_ranges ath10k_sar_freq_ranges[] >> = { >> + { .index = 0, .start_freq = 2412000, .end_freq = 2484000 }, > > 2412 MHz is a center frequency, but other parts of the nl80211 API use > band edges. For example: > > * @NL80211_ATTR_FREQ_RANGE_START: starting frequencry for the > regulatory > * rule in KHz. This is not a center of frequency but an actual > regulatory > * band edge. > * @NL80211_ATTR_FREQ_RANGE_END: ending frequency for the regulatory > rule > * in KHz. This is not a center a frequency but an actual > regulatory > * band edge. > > Seems like we should improve the nl80211.h docs (patch 1) and make > these > edges (this patch). > >> + { .index = 1, .start_freq = 2484000, .end_freq = 5865000 }, >> +}; >> + >> +static const struct cfg80211_sar_capa ath10k_sar_capa = { >> + .type = NL80211_SAR_TYPE_POWER, >> + .num_freq_ranges = (ARRAY_SIZE(ath10k_sar_freq_ranges)), >> + .freq_ranges = &ath10k_sar_freq_ranges[0], >> +}; >> + >> #define ATH10K_MAC_FIRST_OFDM_RATE_IDX 4 >> >> #define ath10k_a_rates (ath10k_rates + >> ATH10K_MAC_FIRST_OFDM_RATE_IDX) >> @@ -2880,6 +2891,95 @@ static int ath10k_mac_vif_recalc_txbf(struct >> ath10k *ar, >> return 0; >> } >> >> +static bool ath10k_mac_is_connected(struct ath10k *ar) >> +{ >> + struct ath10k_vif *arvif; >> + >> + list_for_each_entry(arvif, &ar->arvifs, list) { >> + if (arvif->is_up && arvif->vdev_type == WMI_VDEV_TYPE_STA) >> + return true; >> + } >> + >> + return false; >> +} >> + >> +int ath10k_mac_set_sar_power(struct ath10k *ar) > > This function should be static. > Right. >> +{ >> + int ret; >> + >> + if (!ar->hw_params.dynamic_sar_support) >> + return 0; > > return -EOPNOTSUPP ? > sure >> + >> + if (ar->tx_power_2g_limit == 0 || ar->tx_power_5g_limit == 0) > > ath10k_mac_txpower_recalc() doesn't care about this -- why should you? > This also seems especially weird, because one of the 2 could be valid > nonzero values, and yet you're silently rejecting it. Regardless, the > following seems wrong: > Per current design, it's required for userspace to always set meaningful power limitations. Now in V2, 0 will be treated as "don't have SAR on this range". >> + return 0; > > This should probably be an error. > >> + >> + if (!ath10k_mac_is_connected(ar)) >> + return 0; > > Note to self (since this wasn't obvious to me the first read-through): > you're calling this function from ath10k_bss_assoc() too, so even if > you > weren't connected the first time around, it'll get called later. > >> + >> + ret = ath10k_wmi_pdev_set_param(ar, >> + ar->wmi.pdev_param->txpower_limit2g, >> + ar->tx_power_2g_limit); >> + if (ret) { >> + ath10k_warn(ar, "failed to set 2.4G txpower %d: %d\n", >> + ar->tx_power_2g_limit, ret); >> + return ret; >> + } >> + >> + ret = ath10k_wmi_pdev_set_param(ar, >> + ar->wmi.pdev_param->txpower_limit5g, >> + ar->tx_power_5g_limit); >> + if (ret) { >> + ath10k_warn(ar, "failed to set 5G txpower %d: %d\n", >> + ar->tx_power_5g_limit, ret); >> + return ret; >> + } > > Hmm, so these are the same params configured by > ath10k_mac_txpower_recalc(), except that we're not taking into account > the limitations in ath10k_mac_txpower_recalc() (and vice versa) -- > isn't > that bad? Should we be merging the SAR limitation into > ath10k_mac_txpower_recalc() and calling that instead? > Good suggestions. > Brian > >> + >> + ath10k_dbg(ar, ATH10K_DBG_MAC, "set txpower 2G:%d, 5G:%d >> successfully\n", >> + ar->tx_power_2g_limit, ar->tx_power_5g_limit); >> + >> + return ret; >> +} >> +
On 2020-11-05 07:27, Brian Norris wrote: > Hi Carl, > > Sorry, I lied; I have a few more notes after spending another day > looking at this: > > On Tue, Sep 22, 2020 at 01:49:34PM +0800, Carl Huang wrote: >> --- a/include/net/cfg80211.h >> +++ b/include/net/cfg80211.h >> @@ -1663,6 +1663,55 @@ struct station_info { >> +/** >> + * @struct cfg80211_sar_chan_ranges - sar frequency ranges >> + * @index: the index of this range. It's used to specify >> + * the frequency range when setting SAR power limitation >> + * @start_freq: start channel frequency in kHZ. For example, >> + * 2.4G channel 1 is represented as 2412000 >> + * @end_freq: end channel frequency in kHZ > > If you accept my comments in nl80211.h, you'll want to change this too. > Yes. >> + */ >> +struct cfg80211_sar_freq_ranges { >> + u8 index; >> + u32 start_freq; >> + u32 end_freq; >> +}; > >> --- a/include/uapi/linux/nl80211.h >> +++ b/include/uapi/linux/nl80211.h > >> + * @NL80211_SAR_ATTR_SPECS_START_FREQ: Required (u32) value to >> specify the start >> + * frequency of this range to register SAR capability to wihpy and >> the unit >> + * is kHZ >> + * >> + * @NL80211_SAR_ATTR_SPECS_END_FREQ: Required (u32) value to specify >> the end frequency >> + * of this range to register SAR capability to wiphy and the unit is >> kHZ > > The documentation here isn't clear whether these are center frequencies > or band edges. The cfg80211 comments do though (center freq). However, > this is inconsistent with NL80211_ATTR_FREQ_RANGE_START and > NL80211_ATTR_FREQ_RANGE_END -- I'd suggest being consistent? > Yes. Will change to band edge. >> --- a/net/mac80211/cfg.c >> +++ b/net/mac80211/cfg.c > >> @@ -15331,6 +15496,14 @@ static const struct genl_ops nl80211_ops[] = >> { >> .internal_flags = NL80211_FLAG_NEED_NETDEV | >> NL80211_FLAG_NEED_RTNL, >> }, >> + { >> + .cmd = NL80211_CMD_SET_SAR_SPECS, >> + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, >> + .doit = nl80211_set_sar_specs, >> + .flags = GENL_UNS_ADMIN_PERM, >> + .internal_flags = NL80211_FLAG_NEED_NETDEV_UP | > > Is there a reason this needs to be UP? CMD_SET_WIPHY (which can also > configure TX power, a little differently) does not. Seems like this > could just be NL80211_FLAG_NEED_NETDEV -- or maybe not even that, if we > switch this to a WIPHY command like Johannes noted. > Will change to NL80211_FLAG_NEED_WIPHY. > Brian > >> + NL80211_FLAG_NEED_RTNL, >> + }, >> }; > > _______________________________________________ > ath10k mailing list > ath10k@lists.infradead.org > http://lists.infradead.org/mailman/listinfo/ath10k
On 2020-11-05 01:48, Brian Norris wrote: > On Wed, Nov 4, 2020 at 12:44 AM Carl Huang <cjhuang@codeaurora.org> > wrote: >> On 2020-09-28 20:36, Johannes Berg wrote: >> > On Tue, 2020-09-22 at 13:49 +0800, Carl Huang wrote: >> >> +struct cfg80211_sar_freq_ranges { >> >> + u8 index; >> > >> > Does an index here make sense? >> > >> With agreement from Google, it's OK to remove it. > > I'm not sure "Google" is the arbiter of the nl80211 API, even if we > are the current planned users ;) > > But I think I agree with Johannes, that given the other plans (user > space must send all bands all the time; dropping the "apply to all > bands" support), an index isn't really necessary in either the user > space API or the internal representation handed down to drivers. All > bands should be specified, in order. > > Brian > The index here will be removed. But let's keep the explicit index in SET command. I think it adds no burden to userspace but has flexibility to skip some ranges as we remove "all or nothing" limitation. >> >> + u32 start_freq; >> >> + u32 end_freq; >> >> +};
On Thu, Nov 5, 2020 at 3:27 AM Carl Huang <cjhuang@codeaurora.org> wrote: > On 2020-11-05 07:11, Brian Norris wrote: > > On Tue, Sep 22, 2020 at 01:49:35PM +0800, Carl Huang wrote: > >> + if (ar->tx_power_2g_limit == 0 || ar->tx_power_5g_limit == 0) > > > > ath10k_mac_txpower_recalc() doesn't care about this -- why should you? > > This also seems especially weird, because one of the 2 could be valid > > nonzero values, and yet you're silently rejecting it. Regardless, the > > following seems wrong: > > > Per current design, it's required for userspace to always set meaningful > power limitations. Well, that wasn't specified anywhere (in the docs nor in the nl80211 implementation); if one wanted to enforce that, it should probably go in the nl80211 layer, not the driver implementation. Otherwise, we're very prone to varying driver implementations (not a good thing). > Now in V2, 0 will be treated as "don't have SAR on this range". OK... I guess that's reasonable, although that should be explicitly called out in nl80211.h. Brian
diff --git a/include/net/cfg80211.h b/include/net/cfg80211.h index c9bce9b..94f09dd 100644 --- a/include/net/cfg80211.h +++ b/include/net/cfg80211.h @@ -1663,6 +1663,55 @@ struct station_info { u8 connected_to_as; }; +/** + * struct cfg80211_sar_sub_specs - sub specs limit + * @power: value in 0.25dbm + * @freq_range_index: index the power limitation applies to + */ +struct cfg80211_sar_sub_specs { + u8 power; + u8 freq_range_index; +}; + +/** + * struct cfg80211_sar_specs - sar limit specs + * @type: it's set with power in 0.25dbm or other types + * @num_sub_specs: number of sar sub specs + * @sub_specs: memory to hold the sar sub specs + */ +struct cfg80211_sar_specs { + enum nl80211_sar_type type; + u16 num_sub_specs; + struct cfg80211_sar_sub_specs *sub_specs; +}; + + +/** + * @struct cfg80211_sar_chan_ranges - sar frequency ranges + * @index: the index of this range. It's used to specify + * the frequency range when setting SAR power limitation + * @start_freq: start channel frequency in kHZ. For example, + * 2.4G channel 1 is represented as 2412000 + * @end_freq: end channel frequency in kHZ + */ +struct cfg80211_sar_freq_ranges { + u8 index; + u32 start_freq; + u32 end_freq; +}; + +/** + * struct cfg80211_sar_capa - sar limit capability + * @type: it's set via power in 0.25dbm or other types + * @num_freq_ranges: number of frequency ranges + * @chan_ranges: memory to hold the channel ranges. + */ +struct cfg80211_sar_capa { + enum nl80211_sar_type type; + u8 num_freq_ranges; + const struct cfg80211_sar_freq_ranges *freq_ranges; +}; + #if IS_ENABLED(CONFIG_CFG80211) /** * cfg80211_get_station - retrieve information about a given station @@ -4153,6 +4202,7 @@ struct cfg80211_ops { struct cfg80211_tid_config *tid_conf); int (*reset_tid_config)(struct wiphy *wiphy, struct net_device *dev, const u8 *peer, u8 tids); + int (*set_sar_specs)(struct wiphy *wiphy, struct cfg80211_sar_specs *sar); }; /* @@ -4919,6 +4969,8 @@ struct wiphy { u8 max_data_retry_count; + const struct cfg80211_sar_capa *sar_capa; + char priv[] __aligned(NETDEV_ALIGN); }; diff --git a/include/net/mac80211.h b/include/net/mac80211.h index ec148b3..df758ee 100644 --- a/include/net/mac80211.h +++ b/include/net/mac80211.h @@ -4125,6 +4125,8 @@ struct ieee80211_ops { int (*reset_tid_config)(struct ieee80211_hw *hw, struct ieee80211_vif *vif, struct ieee80211_sta *sta, u8 tids); + int (*set_sar_specs)(struct ieee80211_hw *hw, + const struct cfg80211_sar_specs *sar); }; /** diff --git a/include/uapi/linux/nl80211.h b/include/uapi/linux/nl80211.h index 0584e0d..2dcfed4 100644 --- a/include/uapi/linux/nl80211.h +++ b/include/uapi/linux/nl80211.h @@ -1177,6 +1177,9 @@ * includes the contents of the frame. %NL80211_ATTR_ACK flag is included * if the recipient acknowledged the frame. * + * @NL80211_CMD_SET_SAR_SPECS: SAR power limitation configuration is + * passed using %NL80211_ATTR_SAR_SPEC. + * * @NL80211_CMD_MAX: highest used command number * @__NL80211_CMD_AFTER_LAST: internal use */ @@ -1407,6 +1410,8 @@ enum nl80211_commands { NL80211_CMD_CONTROL_PORT_FRAME_TX_STATUS, + NL80211_CMD_SET_SAR_SPECS, + /* add new commands above here */ /* used to define NL80211_CMD_MAX below */ @@ -2513,6 +2518,10 @@ enum nl80211_commands { * @NL80211_ATTR_HE_6GHZ_CAPABILITY: HE 6 GHz Band Capability element (from * association request when used with NL80211_CMD_NEW_STATION). * + * @NL80211_ATTR_SAR_SPEC: SAR power limitation specification when + * used with %NL80211_CMD_SET_SAR_SPECS. It contains array of + * %nl80211_sar_specs_attrs. + * * @NUM_NL80211_ATTR: total number of nl80211_attrs available * @NL80211_ATTR_MAX: highest attribute number currently defined * @__NL80211_ATTR_AFTER_LAST: internal use @@ -2995,6 +3004,8 @@ enum nl80211_attrs { NL80211_ATTR_HE_6GHZ_CAPABILITY, + NL80211_ATTR_SAR_SPEC, + /* add attributes here, update the policy in nl80211.c */ __NL80211_ATTR_AFTER_LAST, @@ -7004,4 +7015,79 @@ enum nl80211_iftype_akm_attributes { NL80211_IFTYPE_AKM_ATTR_MAX = __NL80211_IFTYPE_AKM_ATTR_LAST - 1, }; +/** + * enum nl80211_sar_type - type of SAR specs + * + * @NL80211_SAR_TYPE_POWER: the value is specified in 0.25 dbm + * + */ +enum nl80211_sar_type { + NL80211_SAR_TYPE_POWER, + + /* add new type here */ + + /* Keep last */ + NUM_NL80211_SAR_TYPE, +}; + +/** + * nl80211_sar_attrs - Attributes for SAR spec + * + * @NL80211_SAR_ATTR_TYPE: Type of SAR: power or index + * + * @NL80211_SAR_ATTR_SPECS: Nested array of SAR power + * limit specifications. Each specification contains a set + * of %nl80211_sar_specs_attrs + * + * @__NL80211_SAR_ATTR_LAST: Internal + * @NL80211_SAR_ATTR_MAX: highest sar attribute + * + * These attributes are used with %NL80211_CMD_SET_SAR_SPEC + */ +enum nl80211_sar_attrs { + __NL80211_SAR_ATTR_INVALID, + + NL80211_SAR_ATTR_TYPE, + NL80211_SAR_ATTR_SPECS, + + __NL80211_SAR_ATTR_LAST, + NL80211_SAR_ATTR_MAX = __NL80211_SAR_ATTR_LAST - 1, +}; + +#define NL80211_SAR_ALL_FREQ_RNAGES 0xff +#define NUM_MAX_NL80211_SAR_FREQ_RANGES 0xfe + +/** + * nl80211_sar_specs_attrs - Attributes for SAR power limit specs + * + * @NL80211_SAR_ATTR_SPECS_POWER: Required (u32)value to specify the actual + * power limit value in units of 0.25 dBm if type is + * NL80211_SAR_TYPE_POWER. (i.e., a value of 44 represents 11 dBm) + * + * @NL80211_SAR_ATTR_SPECS_FREQ_RANGES_INDEX: optional (u32) value to specify the + * index of exported freq ranges table. If this attribute is not present, then + * the power is applied to all freq ranges, i.e, all bands + * + * @NL80211_SAR_ATTR_SPECS_START_FREQ: Required (u32) value to specify the start + * frequency of this range to register SAR capability to wihpy and the unit + * is kHZ + * + * @NL80211_SAR_ATTR_SPECS_END_FREQ: Required (u32) value to specify the end frequency + * of this range to register SAR capability to wiphy and the unit is kHZ + * + * @__NL80211_SAR_ATTR_SPECS_LAST: Internal + * @NL80211_SAR_ATTR_SPECS_MAX: highest sar specs attribute + */ +enum nl80211_sar_specs_attrs { + __NL80211_SAR_ATTR_SPECS_INVALID, + + NL80211_SAR_ATTR_SPECS_POWER, + NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX, + NL80211_SAR_ATTR_SPECS_START_FREQ, + NL80211_SAR_ATTR_SPECS_END_FREQ, + + __NL80211_SAR_ATTR_SPECS_LAST, + NL80211_SAR_ATTR_SPECS_MAX = __NL80211_SAR_ATTR_SPECS_LAST - 1, +}; + #endif /* __LINUX_NL80211_H */ diff --git a/net/mac80211/cfg.c b/net/mac80211/cfg.c index b4e39e3..cc74322 100644 --- a/net/mac80211/cfg.c +++ b/net/mac80211/cfg.c @@ -3993,6 +3993,17 @@ static int ieee80211_reset_tid_config(struct wiphy *wiphy, return ret; } +static int ieee80211_set_sar_specs(struct wiphy *wiphy, + struct cfg80211_sar_specs *sar) +{ + struct ieee80211_local *local = wiphy_priv(wiphy); + + if (!local->ops->set_sar_specs) + return -EOPNOTSUPP; + + return local->ops->set_sar_specs(&local->hw, sar); +} + const struct cfg80211_ops mac80211_config_ops = { .add_virtual_intf = ieee80211_add_iface, .del_virtual_intf = ieee80211_del_iface, @@ -4096,4 +4107,5 @@ const struct cfg80211_ops mac80211_config_ops = { .probe_mesh_link = ieee80211_probe_mesh_link, .set_tid_config = ieee80211_set_tid_config, .reset_tid_config = ieee80211_reset_tid_config, + .set_sar_specs = ieee80211_set_sar_specs, }; diff --git a/net/wireless/nl80211.c b/net/wireless/nl80211.c index 52a35e7..9318b3a 100644 --- a/net/wireless/nl80211.c +++ b/net/wireless/nl80211.c @@ -367,6 +367,19 @@ nl80211_tid_config_attr_policy[NL80211_TID_CONFIG_ATTR_MAX + 1] = { NLA_POLICY_NESTED(nl80211_txattr_policy), }; +static const struct nla_policy +sar_specs_policy[NL80211_SAR_ATTR_SPECS_MAX + 1] = { + [NL80211_SAR_ATTR_SPECS_POWER] = { .type = NLA_U8 }, + [NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX] = + NLA_POLICY_MAX(NLA_U8, NUM_MAX_NL80211_SAR_FREQ_RANGES), +}; + +static const struct nla_policy +sar_policy[NL80211_SAR_ATTR_MAX + 1] = { + [NL80211_SAR_ATTR_TYPE] = NLA_POLICY_MAX(NLA_U32, NUM_NL80211_SAR_TYPE), + [NL80211_SAR_ATTR_SPECS] = NLA_POLICY_NESTED_ARRAY(sar_specs_policy), +}; + static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { [0] = { .strict_start_type = NL80211_ATTR_HE_OBSS_PD }, [NL80211_ATTR_WIPHY] = { .type = NLA_U32 }, @@ -675,6 +688,7 @@ static const struct nla_policy nl80211_policy[NUM_NL80211_ATTR] = { [NL80211_ATTR_SCAN_FREQ_KHZ] = { .type = NLA_NESTED }, [NL80211_ATTR_HE_6GHZ_CAPABILITY] = NLA_POLICY_EXACT_LEN(sizeof(struct ieee80211_he_6ghz_capa)), + [NL80211_ATTR_SAR_SPEC] = NLA_POLICY_NESTED(sar_policy), }; /* policy for the key attributes */ @@ -1856,6 +1870,8 @@ static int nl80211_add_commands_unsplit(struct cfg80211_registered_device *rdev, goto nla_put_failure; } + CMD(set_sar_specs, SET_SAR_SPECS); + return i; nla_put_failure: return -ENOBUFS; @@ -2034,6 +2050,55 @@ nl80211_put_tid_config_support(struct cfg80211_registered_device *rdev, return -ENOBUFS; } +static int +nl80211_put_sar_specs(struct cfg80211_registered_device *rdev, + struct sk_buff *msg) +{ + struct nlattr *sar_capa, *specs, *sub_freq_range; + u8 num_freq_ranges; + int i; + + if (!rdev->wiphy.sar_capa) + return 0; + + num_freq_ranges = rdev->wiphy.sar_capa->num_freq_ranges; + + sar_capa = nla_nest_start(msg, NL80211_ATTR_SAR_SPEC); + if (!sar_capa) + return -ENOSPC; + + if (nla_put_u16(msg, NL80211_SAR_ATTR_TYPE, rdev->wiphy.sar_capa->type)) + goto fail; + + specs = nla_nest_start_noflag(msg, NL80211_SAR_ATTR_SPECS); + if (!specs) + goto fail; + + /* report supported freq_ranges */ + for (i = 0; i < num_freq_ranges; i++) { + sub_freq_range = nla_nest_start_noflag(msg, i + 1); + + nla_put_u32(msg, NL80211_SAR_ATTR_SPECS_START_FREQ, + rdev->wiphy.sar_capa->freq_ranges[i].start_freq); + + nla_put_u32(msg, NL80211_SAR_ATTR_SPECS_END_FREQ, + rdev->wiphy.sar_capa->freq_ranges[i].end_freq); + + nla_put_u8(msg, NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX, + rdev->wiphy.sar_capa->freq_ranges[i].index); + + nla_nest_end(msg, sub_freq_range); + } + + nla_nest_end(msg, specs); + nla_nest_end(msg, sar_capa); + + return 0; +fail: + nla_nest_cancel(msg, sar_capa); + return -ENOBUFS; +} + struct nl80211_dump_wiphy_state { s64 filter_wiphy; long start; @@ -2598,6 +2663,9 @@ static int nl80211_send_wiphy(struct cfg80211_registered_device *rdev, if (nl80211_put_tid_config_support(rdev, msg)) goto nla_put_failure; + if (nl80211_put_sar_specs(rdev, msg)) + goto nla_put_failure; + /* done */ state->split_start = 0; break; @@ -14473,6 +14541,103 @@ static void nl80211_post_doit(const struct genl_ops *ops, struct sk_buff *skb, } } +static int nl80211_set_sar_specs(struct sk_buff *skb, struct genl_info *info) +{ + struct cfg80211_registered_device *rdev = info->user_ptr[0]; + struct nlattr *spec[NL80211_SAR_ATTR_SPECS_MAX + 1]; + struct nlattr *tb[NL80211_SAR_ATTR_MAX + 1]; + struct cfg80211_sar_specs *sar_spec; + u8 type, power, index, specs; + struct nlattr *spec_list; + int rem, err; + + if (!rdev->wiphy.sar_capa) + return -EOPNOTSUPP; + + if (!info->attrs[NL80211_ATTR_SAR_SPEC]) + return -EINVAL; + + nla_parse_nested(tb, NL80211_SAR_ATTR_MAX, info->attrs[NL80211_ATTR_SAR_SPEC], + sar_policy, info->extack); + + if (!tb[NL80211_SAR_ATTR_TYPE]) + return -EINVAL; + + type = nla_get_u32(tb[NL80211_SAR_ATTR_TYPE]); + + if (!tb[NL80211_SAR_ATTR_SPECS]) + return -EINVAL; + + specs = 0; + nla_for_each_nested(spec_list, tb[NL80211_SAR_ATTR_SPECS], rem) + specs++; + + sar_spec = kzalloc(sizeof(*sar_spec) + + specs * sizeof(struct cfg80211_sar_sub_specs), + GFP_KERNEL); + if (!sar_spec) + return -ENOMEM; + + sar_spec->sub_specs = (struct cfg80211_sar_sub_specs *) + ((char *)sar_spec + sizeof(*sar_spec)); + specs = 0; + sar_spec->type = type; + + nla_for_each_nested(spec_list, tb[NL80211_SAR_ATTR_SPECS], rem) { + if (nla_parse(spec, + NL80211_SAR_ATTR_SPECS_MAX, + nla_data(spec_list), + nla_len(spec_list), + sar_specs_policy, + NULL)) { + err = -EINVAL; + goto error; + } + + /* for power type, power value must be presented */ + if (!spec[NL80211_SAR_ATTR_SPECS_POWER] && + type == NL80211_SAR_TYPE_POWER) { + err = -EINVAL; + goto error; + } + + power = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_POWER]); + sar_spec->sub_specs[specs].power = power; + + /* if NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX isn't present, + * then the power applies to all bands. But it's only valid + * for the first entry. + */ + if (!spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]) { + if (specs) { + err = -EINVAL; + goto error; + } else { + sar_spec->sub_specs[specs].freq_range_index = + NL80211_SAR_ALL_FREQ_RNAGES; + specs++; + break; + } + } + + index = nla_get_u8(spec[NL80211_SAR_ATTR_SPECS_FREQ_RANGE_INDEX]); + sar_spec->sub_specs[specs].freq_range_index = index; + specs++; + } + + sar_spec->num_sub_specs = specs; + + rdev->cur_cmd_info = info; + if (rdev->ops->set_sar_specs) + err = rdev_set_sar_specs(rdev, sar_spec); + else + err = -EOPNOTSUPP; + rdev->cur_cmd_info = NULL; +error: + kfree(sar_spec); + return err; +} + static const struct genl_ops nl80211_ops[] = { { .cmd = NL80211_CMD_GET_WIPHY, @@ -15331,6 +15496,14 @@ static const struct genl_ops nl80211_ops[] = { .internal_flags = NL80211_FLAG_NEED_NETDEV | NL80211_FLAG_NEED_RTNL, }, + { + .cmd = NL80211_CMD_SET_SAR_SPECS, + .validate = GENL_DONT_VALIDATE_STRICT | GENL_DONT_VALIDATE_DUMP, + .doit = nl80211_set_sar_specs, + .flags = GENL_UNS_ADMIN_PERM, + .internal_flags = NL80211_FLAG_NEED_NETDEV_UP | + NL80211_FLAG_NEED_RTNL, + }, }; static struct genl_family nl80211_fam __ro_after_init = { diff --git a/net/wireless/rdev-ops.h b/net/wireless/rdev-ops.h index 950d574..d9931c5 100644 --- a/net/wireless/rdev-ops.h +++ b/net/wireless/rdev-ops.h @@ -1356,4 +1356,16 @@ static inline int rdev_reset_tid_config(struct cfg80211_registered_device *rdev, return ret; } +static inline int rdev_set_sar_specs(struct cfg80211_registered_device *rdev, + struct cfg80211_sar_specs *sar) +{ + int ret; + + trace_rdev_set_sar_specs(&rdev->wiphy, sar); + ret = rdev->ops->set_sar_specs(&rdev->wiphy, sar); + trace_rdev_return_int(&rdev->wiphy, ret); + + return ret; +} + #endif /* __CFG80211_RDEV_OPS */ diff --git a/net/wireless/trace.h b/net/wireless/trace.h index 6e218a0..116be64 100644 --- a/net/wireless/trace.h +++ b/net/wireless/trace.h @@ -3547,6 +3547,25 @@ TRACE_EVENT(rdev_reset_tid_config, TP_printk(WIPHY_PR_FMT ", " NETDEV_PR_FMT ", peer: " MAC_PR_FMT ", tids: 0x%x", WIPHY_PR_ARG, NETDEV_PR_ARG, MAC_PR_ARG(peer), __entry->tids) ); + +TRACE_EVENT(rdev_set_sar_specs, + TP_PROTO(struct wiphy *wiphy, struct cfg80211_sar_specs *sar), + TP_ARGS(wiphy, sar), + TP_STRUCT__entry( + WIPHY_ENTRY + __field(u16, type) + __field(u16, num) + ), + TP_fast_assign( + WIPHY_ASSIGN; + __entry->type = sar->type; + __entry->num = sar->num_sub_specs; + + ), + TP_printk(WIPHY_PR_FMT ", Set type:%d, num_specs:%d", + WIPHY_PR_ARG, __entry->type, __entry->num) +); + #endif /* !__RDEV_OPS_TRACE || TRACE_HEADER_MULTI_READ */ #undef TRACE_INCLUDE_PATH
NL80211_CMD_SET_SAR_SPECS is added to configure SAR from user space. NL80211_ATTR_SAR_SPEC is used to pass the SAR power specification when used with NL80211_CMD_SET_SAR_SPECS. Wireless driver needs to register SAR type, supported frequency ranges to wiphy, so user space can query it. The index in frequency range is used to specify which sub band the power limitation applies to. The SAR type is for compatibility, so later other SAR mechanism can be implemented without breaking the user space SAR applications. Normal process is user space quries the SAR capability, and gets the index of supported frequency ranges and associates the power limitation with this index and sends to kernel. Here is an example of message send to kernel: 8c 00 00 00 08 00 03 00 15 00 00 00 38 00 26 81 08 00 01 00 00 00 00 00 2c 00 02 80 14 00 01 80 05 00 02 00 00 00 00 00 05 00 01 00 38 00 00 00 14 00 02 80 05 00 02 00 01 00 00 00 05 00 01 00 48 00 00 00 NL80211_CMD_SET_SAR_SPECS: 0x8c NL80211_ATTR_SAR_SPEC: 0x8126 (NLA_NESTED) NL80211_SAR_ATTR_TYPE: 0x00 (NL80211_SAR_TYPE_POWER) NL80211_SAR_ATTR_SPECS: 0x8002 (NLA_NESTED) freq range 0 power: 0x38 in 0.25dbm unit (14dbm) freq range 1 power: 0x48 in 0.25dbm unit (18dbm) Signed-off-by: Carl Huang <cjhuang@codeaurora.org> --- include/net/cfg80211.h | 52 +++++++++++++ include/net/mac80211.h | 2 + include/uapi/linux/nl80211.h | 86 +++++++++++++++++++++ net/mac80211/cfg.c | 12 +++ net/wireless/nl80211.c | 173 +++++++++++++++++++++++++++++++++++++++++++ net/wireless/rdev-ops.h | 12 +++ net/wireless/trace.h | 19 +++++ 7 files changed, 356 insertions(+)