All,
I have a Seagate Wireless Plus I am trying to put a vanilla kernel on
(currently v3.10-rc5 for various omap DT bits).
I have the board booting and running a debian rootfs on the HD. I just
got mmc to come up and I've discovered that the wireless card is vendor
0x02d0, device 0x4319.
My hope, once I got to this point, was that I would be able to use
the mainline, open source driver. Unfortunately, it looks like brcmfmac
lost support for the 0x4319 while it was in staging.
The commit in question is:
4dad253 staging: brcm80211: remove code for unsupported chip
Is adding it back in a bridge too far?
thx,
Jason.
Added Franky...
On Wed, Jun 12, 2013 at 06:21:23PM -0400, Jason Cooper wrote:
> All,
>
> I have a Seagate Wireless Plus I am trying to put a vanilla kernel on
> (currently v3.10-rc5 for various omap DT bits).
>
> I have the board booting and running a debian rootfs on the HD. I just
> got mmc to come up and I've discovered that the wireless card is vendor
> 0x02d0, device 0x4319.
>
> My hope, once I got to this point, was that I would be able to use
> the mainline, open source driver. Unfortunately, it looks like brcmfmac
> lost support for the 0x4319 while it was in staging.
>
> The commit in question is:
>
> 4dad253 staging: brcm80211: remove code for unsupported chip
>
> Is adding it back in a bridge too far?
It looks like only two things are needed, adding the chip id and then
setting the addresses. Here's a preliminary patch to do just that.
Note, the addresses I have used are a straight copy from the 4329. It
causes this:
[ 12.346466] mmcblk mmc0:0001: no of_node; not parsing pinctrl DT
[ 12.353149] brcmfmac_sdio mmc0:0001:1: no of_node; not parsing pinctrl DT
[ 12.360717] brcmfmac_sdio mmc0:0001:2: no of_node; not parsing pinctrl DT
[ 12.368804] brcmfmac: brcmf_sdio_chip_drivestrengthinit: No SDIO Drive strength init done for chip 4319 rev 1 pmurev 7
[ 12.380462] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
[ 12.391906] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
[ 12.403289] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
[ 12.412261] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
[ 12.418457] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
[ 12.429718] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
[ 12.441101] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
[ 12.450073] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
I've taken a quick look at aosp, the provided GPL broadcom code, and I
haven't seen anything for the 4319 wrt to addresses. How hard would it
be to get those addresses from you guys?
thx,
Jason.
----8<------
commit 8fa440e8d6da21c95cdbc9c1ced76d5962cd9e3d
Author: Jason Cooper <[email protected]>
Date: Mon Jun 17 15:30:43 2013 +0000
net: brcmfmac: add sdio chip id 0x4319
Signed-off-by: Jason Cooper <[email protected]>
diff --git a/drivers/net/wireless/brcm80211/brcmfmac/bcmsdh_sdmmc.c b/drivers/net/wireless/brcm80211/brcmfmac/bcmsdh_sdmmc.c
index 44fa0cd..75ddc2b 100644
--- a/drivers/net/wireless/brcm80211/brcmfmac/bcmsdh_sdmmc.c
+++ b/drivers/net/wireless/brcm80211/brcmfmac/bcmsdh_sdmmc.c
@@ -43,6 +43,7 @@
#define SDIO_DEVICE_ID_BROADCOM_43143 43143
#define SDIO_DEVICE_ID_BROADCOM_43241 0x4324
+#define SDIO_DEVICE_ID_BROADCOM_4319 0x4319
#define SDIO_DEVICE_ID_BROADCOM_4329 0x4329
#define SDIO_DEVICE_ID_BROADCOM_4330 0x4330
#define SDIO_DEVICE_ID_BROADCOM_4334 0x4334
@@ -55,6 +56,7 @@
static const struct sdio_device_id brcmf_sdmmc_ids[] = {
{SDIO_DEVICE(SDIO_VENDOR_ID_BROADCOM, SDIO_DEVICE_ID_BROADCOM_43143)},
{SDIO_DEVICE(SDIO_VENDOR_ID_BROADCOM, SDIO_DEVICE_ID_BROADCOM_43241)},
+ {SDIO_DEVICE(SDIO_VENDOR_ID_BROADCOM, SDIO_DEVICE_ID_BROADCOM_4319)},
{SDIO_DEVICE(SDIO_VENDOR_ID_BROADCOM, SDIO_DEVICE_ID_BROADCOM_4329)},
{SDIO_DEVICE(SDIO_VENDOR_ID_BROADCOM, SDIO_DEVICE_ID_BROADCOM_4330)},
{SDIO_DEVICE(SDIO_VENDOR_ID_BROADCOM, SDIO_DEVICE_ID_BROADCOM_4334)},
diff --git a/drivers/net/wireless/brcm80211/brcmfmac/dhd_sdio.c b/drivers/net/wireless/brcm80211/brcmfmac/dhd_sdio.c
index d248751..68816ec 100644
--- a/drivers/net/wireless/brcm80211/brcmfmac/dhd_sdio.c
+++ b/drivers/net/wireless/brcm80211/brcmfmac/dhd_sdio.c
@@ -3551,6 +3551,8 @@ static bool brcmf_sdbrcm_chipmatch(u16 chipid)
return true;
if (chipid == BCM43241_CHIP_ID)
return true;
+ if (chipid == BCM4319_CHIP_ID)
+ return true;
if (chipid == BCM4329_CHIP_ID)
return true;
if (chipid == BCM4330_CHIP_ID)
diff --git a/drivers/net/wireless/brcm80211/brcmfmac/sdio_chip.c b/drivers/net/wireless/brcm80211/brcmfmac/sdio_chip.c
index ca72177..4336dbb 100644
--- a/drivers/net/wireless/brcm80211/brcmfmac/sdio_chip.c
+++ b/drivers/net/wireless/brcm80211/brcmfmac/sdio_chip.c
@@ -31,6 +31,15 @@
#include "sdio_chip.h"
/* chip core base & ramsize */
+/* bcm4319 */
+/* SDIO device core, ID 0x829 */
+#define BCM4319_CORE_BUS_BASE 0x18011000
+/* internal memory core, ID 0x80e */
+#define BCM4319_CORE_SOCRAM_BASE 0x18003000
+/* ARM Cortex M3 core, ID 0x82a */
+#define BCM4319_CORE_ARM_BASE 0x18002000
+#define BCM4319_RAMSIZE 0x48000
+
/* bcm4329 */
/* SDIO device core, ID 0x829 */
#define BCM4329_CORE_BUS_BASE 0x18011000
@@ -484,6 +493,15 @@ static int brcmf_sdio_chip_recognition(struct brcmf_sdio_dev *sdiodev,
ci->c_inf[3].cib = 0x07004211;
ci->ramsize = 0x90000;
break;
+ case BCM4319_CHIP_ID:
+ ci->c_inf[1].id = BCMA_CORE_SDIO_DEV;
+ ci->c_inf[1].base = BCM4319_CORE_BUS_BASE;
+ ci->c_inf[2].id = BCMA_CORE_INTERNAL_MEM;
+ ci->c_inf[2].base = BCM4319_CORE_SOCRAM_BASE;
+ ci->c_inf[3].id = BCMA_CORE_ARM_CM3;
+ ci->c_inf[3].base = BCM4319_CORE_ARM_BASE;
+ ci->ramsize = BCM4319_RAMSIZE;
+ break;
case BCM4329_CHIP_ID:
ci->c_inf[1].id = BCMA_CORE_SDIO_DEV;
ci->c_inf[1].base = BCM4329_CORE_BUS_BASE;
diff --git a/drivers/net/wireless/brcm80211/include/brcm_hw_ids.h b/drivers/net/wireless/brcm80211/include/brcm_hw_ids.h
index c1fe245..c2434db 100644
--- a/drivers/net/wireless/brcm80211/include/brcm_hw_ids.h
+++ b/drivers/net/wireless/brcm80211/include/brcm_hw_ids.h
@@ -36,6 +36,7 @@
#define BCM43236_CHIP_ID 43236
#define BCM43238_CHIP_ID 43238
#define BCM43241_CHIP_ID 0x4324
+#define BCM4319_CHIP_ID 0x4319
#define BCM4329_CHIP_ID 0x4329
#define BCM4330_CHIP_ID 0x4330
#define BCM4331_CHIP_ID 0x4331
2013/6/17 Franky Lin <[email protected]>:
> On 06/17/2013 09:19 AM, Jason Cooper wrote:
>>
>> Added Franky...
>>
>> On Wed, Jun 12, 2013 at 06:21:23PM -0400, Jason Cooper wrote:
>>>
>>> All,
>>>
>>> I have a Seagate Wireless Plus I am trying to put a vanilla kernel on
>>> (currently v3.10-rc5 for various omap DT bits).
>>>
>>> I have the board booting and running a debian rootfs on the HD. I just
>>> got mmc to come up and I've discovered that the wireless card is vendor
>>> 0x02d0, device 0x4319.
>>>
>>> My hope, once I got to this point, was that I would be able to use
>>> the mainline, open source driver. Unfortunately, it looks like brcmfmac
>>> lost support for the 0x4319 while it was in staging.
>>>
>>> The commit in question is:
>>>
>>> 4dad253 staging: brcm80211: remove code for unsupported chip
>>>
>>> Is adding it back in a bridge too far?
>>
>> It looks like only two things are needed, adding the chip id and then
>> setting the addresses. Here's a preliminary patch to do just that.
>> Note, the addresses I have used are a straight copy from the 4329. It
>> causes this:
>>
>> [ 12.346466] mmcblk mmc0:0001: no of_node; not parsing pinctrl DT
>> [ 12.353149] brcmfmac_sdio mmc0:0001:1: no of_node; not parsing pinctrl
>> DT
>> [ 12.360717] brcmfmac_sdio mmc0:0001:2: no of_node; not parsing pinctrl
>> DT
>> [ 12.368804] brcmfmac: brcmf_sdio_chip_drivestrengthinit: No SDIO Drive
>> strength init done for chip 4319 rev 1 pmurev 7
>> [ 12.380462] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
>> F0:@0x00408=03, Err: -22
>> [ 12.391906] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
>> F0:@0x00408=03, Err: -22
>> [ 12.403289] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
>> F0:@0x00408=03, Err: -22
>> [ 12.412261] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
>> [ 12.418457] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
>> F0:@0x00408=01, Err: -22
>> [ 12.429718] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
>> F0:@0x00408=01, Err: -22
>> [ 12.441101] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
>> F0:@0x00408=01, Err: -22
>> [ 12.450073] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
>>
>> I've taken a quick look at aosp, the provided GPL broadcom code, and I
>> haven't seen anything for the 4319 wrt to addresses. How hard would it
>> be to get those addresses from you guys?
>
>
> Hi Jason,
>
> The major obstacle of adding 4319 support is the obsolete firmware. Since
> it's an EOL chip we are not planning to add the support to brcmfmac.
Does it mean, BCM4319 is a "new architecture" chipset, working
similarly to the already supported (in brcmfmac) chipsets?
When thinking about "old architecture" I mean BCM43236, which AFAIR
requires some different driver architecture with some extra memory
sharing or sth.
--
Rafał
On 06/17/2013 09:19 AM, Jason Cooper wrote:
> Added Franky...
>
> On Wed, Jun 12, 2013 at 06:21:23PM -0400, Jason Cooper wrote:
>> All,
>>
>> I have a Seagate Wireless Plus I am trying to put a vanilla kernel on
>> (currently v3.10-rc5 for various omap DT bits).
>>
>> I have the board booting and running a debian rootfs on the HD. I just
>> got mmc to come up and I've discovered that the wireless card is vendor
>> 0x02d0, device 0x4319.
>>
>> My hope, once I got to this point, was that I would be able to use
>> the mainline, open source driver. Unfortunately, it looks like brcmfmac
>> lost support for the 0x4319 while it was in staging.
>>
>> The commit in question is:
>>
>> 4dad253 staging: brcm80211: remove code for unsupported chip
>>
>> Is adding it back in a bridge too far?
> It looks like only two things are needed, adding the chip id and then
> setting the addresses. Here's a preliminary patch to do just that.
> Note, the addresses I have used are a straight copy from the 4329. It
> causes this:
>
> [ 12.346466] mmcblk mmc0:0001: no of_node; not parsing pinctrl DT
> [ 12.353149] brcmfmac_sdio mmc0:0001:1: no of_node; not parsing pinctrl DT
> [ 12.360717] brcmfmac_sdio mmc0:0001:2: no of_node; not parsing pinctrl DT
> [ 12.368804] brcmfmac: brcmf_sdio_chip_drivestrengthinit: No SDIO Drive strength init done for chip 4319 rev 1 pmurev 7
> [ 12.380462] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
> [ 12.391906] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
> [ 12.403289] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
> [ 12.412261] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
> [ 12.418457] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
> [ 12.429718] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
> [ 12.441101] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
> [ 12.450073] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
>
> I've taken a quick look at aosp, the provided GPL broadcom code, and I
> haven't seen anything for the 4319 wrt to addresses. How hard would it
> be to get those addresses from you guys?
Hi Jason,
The major obstacle of adding 4319 support is the obsolete firmware.
Since it's an EOL chip we are not planning to add the support to brcmfmac.
Franky
On Mon, Jun 17, 2013 at 07:45:25PM +0200, Rafał Miłecki wrote:
> 2013/6/17 Franky Lin <[email protected]>:
> > On 06/17/2013 09:19 AM, Jason Cooper wrote:
> >>
> >> Added Franky...
> >>
> >> On Wed, Jun 12, 2013 at 06:21:23PM -0400, Jason Cooper wrote:
> >>>
> >>> All,
> >>>
> >>> I have a Seagate Wireless Plus I am trying to put a vanilla kernel on
> >>> (currently v3.10-rc5 for various omap DT bits).
> >>>
> >>> I have the board booting and running a debian rootfs on the HD. I just
> >>> got mmc to come up and I've discovered that the wireless card is vendor
> >>> 0x02d0, device 0x4319.
> >>>
> >>> My hope, once I got to this point, was that I would be able to use
> >>> the mainline, open source driver. Unfortunately, it looks like brcmfmac
> >>> lost support for the 0x4319 while it was in staging.
> >>>
> >>> The commit in question is:
> >>>
> >>> 4dad253 staging: brcm80211: remove code for unsupported chip
> >>>
> >>> Is adding it back in a bridge too far?
> >>
> >> It looks like only two things are needed, adding the chip id and then
> >> setting the addresses. Here's a preliminary patch to do just that.
> >> Note, the addresses I have used are a straight copy from the 4329. It
> >> causes this:
> >>
> >> [ 12.346466] mmcblk mmc0:0001: no of_node; not parsing pinctrl DT
> >> [ 12.353149] brcmfmac_sdio mmc0:0001:1: no of_node; not parsing pinctrl
> >> DT
> >> [ 12.360717] brcmfmac_sdio mmc0:0001:2: no of_node; not parsing pinctrl
> >> DT
> >> [ 12.368804] brcmfmac: brcmf_sdio_chip_drivestrengthinit: No SDIO Drive
> >> strength init done for chip 4319 rev 1 pmurev 7
> >> [ 12.380462] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
> >> F0:@0x00408=03, Err: -22
> >> [ 12.391906] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
> >> F0:@0x00408=03, Err: -22
> >> [ 12.403289] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
> >> F0:@0x00408=03, Err: -22
> >> [ 12.412261] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
> >> [ 12.418457] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
> >> F0:@0x00408=01, Err: -22
> >> [ 12.429718] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
> >> F0:@0x00408=01, Err: -22
> >> [ 12.441101] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte
> >> F0:@0x00408=01, Err: -22
> >> [ 12.450073] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
> >>
> >> I've taken a quick look at aosp, the provided GPL broadcom code, and I
> >> haven't seen anything for the 4319 wrt to addresses. How hard would it
> >> be to get those addresses from you guys?
> >
> >
> > Hi Jason,
> >
> > The major obstacle of adding 4319 support is the obsolete firmware. Since
> > it's an EOL chip we are not planning to add the support to brcmfmac.
>
> Does it mean, BCM4319 is a "new architecture" chipset, working
> similarly to the already supported (in brcmfmac) chipsets?
In older versions of the broadcom driver 4329 and 4319 were in the same
code base. There was minimal reference to 4319 beyond "I work with this
chip id". Hence my attempt.
Now I'm wondering how much effort there would be to adding legacy
firmware support to brcmfmac. I would only need AP mode to start with
(yeah, no big deal ;-) ).
thx,
Jason.
On Mon, Jun 17, 2013 at 10:38:23AM -0700, Franky Lin wrote:
> On 06/17/2013 09:19 AM, Jason Cooper wrote:
> >Added Franky...
> >
> >On Wed, Jun 12, 2013 at 06:21:23PM -0400, Jason Cooper wrote:
> >>All,
> >>
> >>I have a Seagate Wireless Plus I am trying to put a vanilla kernel on
> >>(currently v3.10-rc5 for various omap DT bits).
> >>
> >>I have the board booting and running a debian rootfs on the HD. I just
> >>got mmc to come up and I've discovered that the wireless card is vendor
> >>0x02d0, device 0x4319.
> >>
> >>My hope, once I got to this point, was that I would be able to use
> >>the mainline, open source driver. Unfortunately, it looks like brcmfmac
> >>lost support for the 0x4319 while it was in staging.
> >>
> >>The commit in question is:
> >>
> >> 4dad253 staging: brcm80211: remove code for unsupported chip
> >>
> >>Is adding it back in a bridge too far?
> >It looks like only two things are needed, adding the chip id and then
> >setting the addresses. Here's a preliminary patch to do just that.
> >Note, the addresses I have used are a straight copy from the 4329. It
> >causes this:
> >
> >[ 12.346466] mmcblk mmc0:0001: no of_node; not parsing pinctrl DT
> >[ 12.353149] brcmfmac_sdio mmc0:0001:1: no of_node; not parsing pinctrl DT
> >[ 12.360717] brcmfmac_sdio mmc0:0001:2: no of_node; not parsing pinctrl DT
> >[ 12.368804] brcmfmac: brcmf_sdio_chip_drivestrengthinit: No SDIO Drive strength init done for chip 4319 rev 1 pmurev 7
> >[ 12.380462] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
> >[ 12.391906] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
> >[ 12.403289] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=03, Err: -22
> >[ 12.412261] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
> >[ 12.418457] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
> >[ 12.429718] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
> >[ 12.441101] brcmfmac: brcmf_sdioh_request_byte: Failed to write byte F0:@0x00408=01, Err: -22
> >[ 12.450073] brcmfmac: brcmf_sdio_regrw_helper: failed with -22
> >
> >I've taken a quick look at aosp, the provided GPL broadcom code, and I
> >haven't seen anything for the 4319 wrt to addresses. How hard would it
> >be to get those addresses from you guys?
>
> Hi Jason,
>
> The major obstacle of adding 4319 support is the obsolete firmware.
> Since it's an EOL chip we are not planning to add the support to
> brcmfmac.
Let me guess, the firmware would have to be updated to work with
brcmfmac? What about adding a BRMCFMAC_LEGACY option to allow
interfacing with the old firmware?
It may be EOL, but new products are still shipping with this chip
(Wireless Plus, other Azurewave products).
thx,
Jason.