Re: [PATCH 07/34] brcmfmac: Remove brcmf_sdiod_request_data()

2017-08-30 Thread Arend van Spriel
On 19-08-17 22:02, Ian Molton wrote: On 07/08/17 18:51, Ian Molton wrote: On 07/08/17 12:25, Arend van Spriel wrote: Handling of -ENOMEDIUM is altered, but as that's pretty much broken anyway we can ignore that. Please explain why you think it is broken. Not got the code to hand right now,

Re: [PATCH 07/34] brcmfmac: Remove brcmf_sdiod_request_data()

2017-08-19 Thread Ian Molton
On 07/08/17 18:51, Ian Molton wrote: > On 07/08/17 12:25, Arend van Spriel wrote: >>> Handling of -ENOMEDIUM is altered, but as that's pretty much broken >>> anyway >>> we can ignore that. >> >> Please explain why you think it is broken. > > Not got the code to hand right now, but from memory, the

Re: [PATCH 07/34] brcmfmac: Remove brcmf_sdiod_request_data()

2017-08-07 Thread Ian Molton
On 07/08/17 12:25, Arend van Spriel wrote: >> Handling of -ENOMEDIUM is altered, but as that's pretty much broken >> anyway >> we can ignore that. > > Please explain why you think it is broken. Not got the code to hand right now, but from memory, theres a trapdoor case where the state can wind up

Re: [PATCH 07/34] brcmfmac: Remove brcmf_sdiod_request_data()

2017-08-07 Thread Arend van Spriel
On 26-07-17 22:25, Ian Molton wrote: This function is obfuscating how IO works on this chip. Remove it and push its logic into brcmf_sdiod_reg_{read,write}(). Handling of -ENOMEDIUM is altered, but as that's pretty much broken anyway we can ignore that. Please explain why you think it is broke

[PATCH 07/34] brcmfmac: Remove brcmf_sdiod_request_data()

2017-07-26 Thread Ian Molton
This function is obfuscating how IO works on this chip. Remove it and push its logic into brcmf_sdiod_reg_{read,write}(). Handling of -ENOMEDIUM is altered, but as that's pretty much broken anyway we can ignore that. Signed-off-by: Ian Molton # Conflicts: # drivers/net/wireless/broadcom/b

[PATCH 07/34] brcmfmac: Remove brcmf_sdiod_request_data()

2017-07-19 Thread Ian Molton
This function is obfuscating how IO works on this chip. Remove it and push its logic into brcmf_sdiod_reg_{read,write}(). Handling of -ENOMEDIUM is altered, but as that's pretty much broken anyway we can ignore that. Signed-off-by: Ian Molton --- .../wireless/broadcom/brcm80211/brcmfmac/bcmsdh.