Messages in this thread | | | From | Ping-Ke Shih <> | Subject | Re: [PATCH 1/4] rtw88: Add packed attribute to the eFuse structs | Date | Thu, 29 Dec 2022 11:35:59 +0000 |
| |
On Thu, 2022-12-29 at 11:37 +0100, Martin Blumenstingl wrote: > Hi Ping-Ke, > > On Thu, Dec 29, 2022 at 10:25 AM Ping-Ke Shih <pkshih@realtek.com> wrote: > [...] > > > @@ -43,13 +43,13 @@ struct rtw8821ce_efuse { > > > u8 link_cap[4]; > > > u8 link_control[2]; > > > u8 serial_number[8]; > > > - u8 res0:2; /* 0xf4 */ > > > - u8 ltr_en:1; > > > - u8 res1:2; > > > - u8 obff:2; > > > - u8 res2:3; > > > - u8 obff_cap:2; > > > - u8 res3:4; > > > + u16 res0:2; /* 0xf4 */ > > > + u16 ltr_en:1; > > > + u16 res1:2; > > > + u16 obff:2; > > > + u16 res2:3; > > > + u16 obff_cap:2; > > > + u16 res3:4; > > > > These should be __le16. Though bit fields are suitable to efuse layout, > > we don't access these fields for now. It would be well. > My understanding is that it should look like this (replacing all of res0..res3): > __le16 some_field_name; /* 0xf4 */ > How to call that single __le16 field then?
You are right. Maybe, we can name it 'pcie_cap'. But, we don't use them for now, so it is harmless to preserve them as is.
> > I also tried using bit-fields for an __le16 (so basically the same as > my patch but using __le16 instead of u16) but that makes sparse > complain: > error: invalid bitfield specifier for type restricted __le16 > >
We can fix it by:
u8 res0:2; /* 0xf4 */ u8 ltr_en:1; u8 res1:2; u8 obff:2; - u8 res2:3; + u8 res2_1:1; + u8 res2_2:2; u8 obff_cap:2; u8 res3:4;
I'm not sure why people merge bit fields res2_1:1 and res2_2:2 that should be in different u8. I have confirmed this with internal data.
-- Ping-Ke
| |