Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
On 14/06/2014 23:08, José Vázquez wrote: The main problem with the wifi in the Lantiq target are the ARV boards: a lot of people, John included, spent a lot of time an still there are some problems, as you see. bollocks .. i never spent any time on that code. i don't even have the boards so i cannot test which is why its most likely why its broken. if people had spend that much time on the code it would be woring and not ugly and broken. as for the patches you sent, they were not merged because they are a big pile of copy pasta churn that adds lots of redunancy and duplication. it would make the code even uglier than it is now. ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
2014-06-15 8:03 GMT+02:00, John Crispin j...@phrozen.org: On 14/06/2014 23:08, José Vázquez wrote: The main problem with the wifi in the Lantiq target are the ARV boards: a lot of people, John included, spent a lot of time an still there are some problems, as you see. bollocks .. i never spent any time on that code. i don't even have the boards so i cannot test which is why its most likely why its broken. if people had spend that much time on the code it would be woring and not ugly and broken. as for the patches you sent, they were not merged because they are a big pile of copy pasta churn that adds lots of redundancy and duplication. it would make the code even uglier than it is now. You are right: those patches had a lot of redundant code. I will take more care next time, but the original code was not mine. I have a question flying around my head: why nobody realized, apart of you, the potential problems of those patches and the redundant code, and told something to improve them? My apologies: José Vázquez ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
[OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
Hi, Since changeset 40948, calibration data is no longer correctly loaded from EEPROM on the BTHOMEHUBV2B. I've been trying to make sense of the various changes to 0010-MIPS-lantiq-wifi-and-ethernet-eeprom-handling.patch and it looks to me as though merging José's recent patch for the ARV4518PW has left us with an old version of the ath9k EEPROM loading code, without the succession of changes made last year culminating in patch #4417 from Daniel Gimpelevitch. I should imagine therefore that the DGN3500 is broken as well as the BTHOMEHUBV2B. Ben ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
On 14/06/2014 13:29, Ben Mulvihill wrote: Hi, Since changeset 40948, calibration data is no longer correctly loaded from EEPROM on the BTHOMEHUBV2B. I've been trying to make sense of the various changes to 0010-MIPS-lantiq-wifi-and-ethernet-eeprom-handling.patch and it looks to me as though merging José's recent patch for the ARV4518PW has left us with an old version of the ath9k EEPROM loading code, without the succession of changes made last year culminating in patch #4417 from Daniel Gimpelevitch. I should imagine therefore that the DGN3500 is broken as well as the BTHOMEHUBV2B. Ben ok, i will look into it and merge the version we had previous to the 3.10 rebase John u ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
On Sat, 2014-06-14 at 14:06 +0200, John Crispin wrote: On 14/06/2014 13:29, Ben Mulvihill wrote: Hi, Since changeset 40948, calibration data is no longer correctly loaded from EEPROM on the BTHOMEHUBV2B. I've been trying to make sense of the various changes to 0010-MIPS-lantiq-wifi-and-ethernet-eeprom-handling.patch and it looks to me as though merging José's recent patch for the ARV4518PW has left us with an old version of the ath9k EEPROM loading code, without the succession of changes made last year culminating in patch #4417 from Daniel Gimpelevitch. I should imagine therefore that the DGN3500 is broken as well as the BTHOMEHUBV2B. Ben ok, i will look into it and merge the version we had previous to the 3.10 rebase John u Thank you. Let me know if you want me to test anything, or if there is anything else I can do. ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
2014-06-14 14:32 GMT+02:00, Ben Mulvihill ben.mulvih...@gmail.com: On Sat, 2014-06-14 at 14:06 +0200, John Crispin wrote: On 14/06/2014 13:29, Ben Mulvihill wrote: Hi, Since changeset 40948, calibration data is no longer correctly loaded from EEPROM on the BTHOMEHUBV2B. I've been trying to make sense of the various changes to 0010-MIPS-lantiq-wifi-and-ethernet-eeprom-handling.patch and it looks to me as though merging José's recent patch for the ARV4518PW has left us with an old version of the ath9k EEPROM loading code, without the succession of changes made last year culminating in patch #4417 from Daniel Gimpelevitch. I should imagine therefore that the DGN3500 is broken as well as the BTHOMEHUBV2B. Ben ok, i will look into it and merge the version we had previous to the 3.10 rebase John u Thank you. Let me know if you want me to test anything, or if there is anything else I can do. Unless the BTHOMEHUBV2B has an Atheros b/g wireless chip the patch should have no effect in the routers that need ath9k driver. My main concern with that patch were the Siemens SX76x but not the others. AFAIK the cal_data partition is a bit problematic because each manufacturer make the things in its own way, and Daniel Gimpelevitch and Álvaro Fernández know how to fix this problem in the less traumatic way. Regards: José Vázquez ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
On Sat, 2014-06-14 at 18:53 +0200, José Vázquez wrote: 2014-06-14 14:32 GMT+02:00, Ben Mulvihill ben.mulvih...@gmail.com: On Sat, 2014-06-14 at 14:06 +0200, John Crispin wrote: On 14/06/2014 13:29, Ben Mulvihill wrote: Hi, Since changeset 40948, calibration data is no longer correctly loaded from EEPROM on the BTHOMEHUBV2B. I've been trying to make sense of the various changes to 0010-MIPS-lantiq-wifi-and-ethernet-eeprom-handling.patch and it looks to me as though merging José's recent patch for the ARV4518PW has left us with an old version of the ath9k EEPROM loading code, without the succession of changes made last year culminating in patch #4417 from Daniel Gimpelevitch. I should imagine therefore that the DGN3500 is broken as well as the BTHOMEHUBV2B. Ben ok, i will look into it and merge the version we had previous to the 3.10 rebase John u Thank you. Let me know if you want me to test anything, or if there is anything else I can do. Unless the BTHOMEHUBV2B has an Atheros b/g wireless chip the patch should have no effect in the routers that need ath9k driver. My main concern with that patch were the Siemens SX76x but not the others. AFAIK the cal_data partition is a bit problematic because each manufacturer make the things in its own way, and Daniel Gimpelevitch and Álvaro Fernández know how to fix this problem in the less traumatic way. Regards: José Vázquez I've just had a closer look, and realised that only one of your recent patches (http://patchwork.openwrt.org/patch/5582/) has actually been applied, in changeset 40999. As you say, that shouldn't cause problems. The BTHOMEHUBV2B is definitely broken though. and what has broken things is the previous changeset, 40948. Let's see what John comes up with. Your other two patches, #5453 and #5454, are marked as non-applicable in patchwork. Do you still need them? Regards, Ben ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
2014-06-14 21:47 GMT+02:00, Ben Mulvihill ben.mulvih...@gmail.com: On Sat, 2014-06-14 at 18:53 +0200, José Vázquez wrote: Unless the BTHOMEHUBV2B has an Atheros b/g wireless chip the patch should have no effect in the routers that need ath9k driver. My main concern with that patch were the Siemens SX76x but not the others. AFAIK the cal_data partition is a bit problematic because each manufacturer make the things in its own way, and Daniel Gimpelevitch and Álvaro Fernández know how to fix this problem in the less traumatic way. Regards: José Vázquez I've just had a closer look, and realised that only one of your recent patches (http://patchwork.openwrt.org/patch/5582/) has actually been applied, in changeset 40999. As you say, that shouldn't cause problems. The BTHOMEHUBV2B is definitely broken though. and what has broken things is the previous changeset, 40948. Let's see what John comes up with. Your other two patches, #5453 and #5454, are marked as non-applicable in patchwork. Do you still need them? Regards, Ben Patches 5453 and 5454 were marked as not applicable due to the changeset 40948. Were an attemp to correct the problems reading cal_data in Arcadyan/Astoria boards and were wrote by two seguridadwireless forum users. They wrote the code trying to make it backwards compatible. Feel free to use them. The main problem with the wifi in the Lantiq target are the ARV boards: a lot of people, John included, spent a lot of time an still there are some problems, as you see. Regards: Pepe ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel
Re: [OpenWrt-Devel] changeset 40948 breaks loading ath9k calibration data from EEPROM
On Sat, 2014-06-14 at 23:08 +0200, José Vázquez wrote: 2014-06-14 21:47 GMT+02:00, Ben Mulvihill ben.mulvih...@gmail.com: On Sat, 2014-06-14 at 18:53 +0200, José Vázquez wrote: Unless the BTHOMEHUBV2B has an Atheros b/g wireless chip the patch should have no effect in the routers that need ath9k driver. My main concern with that patch were the Siemens SX76x but not the others. AFAIK the cal_data partition is a bit problematic because each manufacturer make the things in its own way, and Daniel Gimpelevitch and Álvaro Fernández know how to fix this problem in the less traumatic way. Regards: José Vázquez I've just had a closer look, and realised that only one of your recent patches (http://patchwork.openwrt.org/patch/5582/) has actually been applied, in changeset 40999. As you say, that shouldn't cause problems. The BTHOMEHUBV2B is definitely broken though. and what has broken things is the previous changeset, 40948. Let's see what John comes up with. Your other two patches, #5453 and #5454, are marked as non-applicable in patchwork. Do you still need them? Regards, Ben Patches 5453 and 5454 were marked as not applicable due to the changeset 40948. Were an attemp to correct the problems reading cal_data in Arcadyan/Astoria boards and were wrote by two seguridadwireless forum users. They wrote the code trying to make it backwards compatible. Feel free to use them. The main problem with the wifi in the Lantiq target are the ARV boards: a lot of people, John included, spent a lot of time an still there are some problems, as you see. Regards: Pepe Thanks for clarifying the origin of patches 5453 and 5454. I can see that following changeset 40948 they will not apply in their current form. What I really wanted to know was whether the combination of changesets 40948 and 40999 has solved the problems on the ARV boards that those two patches were intended to solve. Thanks, Ben ___ openwrt-devel mailing list openwrt-devel@lists.openwrt.org https://lists.openwrt.org/cgi-bin/mailman/listinfo/openwrt-devel