powerpc/5200: make BestComm gen_bd microcode exchangeable

This patch adds a method for defining different microcodes than the pe-defi= ned ones for the MPC52xx's BestComm General Buffer Descriptor (gen_db) task= s. The default microcode is still the one from bcom_gen_bd_[rt]x_task, but=
it can be replaced by calling bcom_gen_bd_set_microcode() which is more ef=
ficient than explicitly loading it via bcom_load_image() after each bcom_ge= n_bd_[rt]x_reset().

Except for a fixed tab vs. space formatting and the transfer format, there = are no diffierences to the initial attempt.

---

diff -uprN -X linux-2.6.32-rc3/Documentation/dontdiff linux-2.6.32-rc3-orig= /arch/powerpc/sysdev/bestcomm/gen_bd.c linux-2.6.32-rc3/arch/powerpc/sysdev= /bestcomm/gen_bd.c
--- linux-2.6.32-rc3-orig/arch/powerpc/sysdev/bestcomm/gen_bd.c 2009-10-05 =
02:12:30.000000000 +0200
+++ linux-2.6.32-rc3/arch/powerpc/sysdev/bestcomm/gen_bd.c 2009-10-05 14:52=
:31.000000000 +0200
@@ -78,6 +78,7 @@ struct bcom_gen_bd_priv {
int initiator;
int ipr;
int maxbufsize;
+ u32 *microcode;
};
=20
=20
@@ -104,6 +105,7 @@ bcom_gen_bd_rx_init(int queue_len, phys_
priv->initiator =3D initiator;
priv->ipr =3D ipr;
priv->maxbufsize =3D maxbufsize;
+ priv->microcode =3D bcom_gen_bd_rx_task;
=20
if (bcom_gen_bd_rx_reset(tsk)) {
bcom_task_free(tsk);
@@ -128,7 +130,7 @@ bcom_gen_bd_rx_reset(struct bcom_task *t
var =3D (struct bcom_gen_bd_rx_var *) bcom_task_var(tsk->tasknum); inc =3D (struct bcom_gen_bd_rx_inc *) bcom_task_inc(tsk->tasknum);
=20
- if (bcom_load_image(tsk->tasknum, bcom_gen_bd_rx_task))
+ if (bcom_load_image(tsk->tasknum, priv->microcode))
return -1;
=20
var->enable =3D bcom_eng->regs_base +
@@ -188,6 +190,7 @@ bcom_gen_bd_tx_init(int queue_len, phys_
priv->fifo =3D fifo;
priv->initiator =3D initiator;
priv->ipr =3D ipr;
+ priv->microcode =3D bcom_gen_bd_tx_task;
=20
if (bcom_gen_bd_tx_reset(tsk)) {
bcom_task_free(tsk);
@@ -212,7 +215,7 @@ bcom_gen_bd_tx_reset(struct bcom_task *t
var =3D (struct bcom_gen_bd_tx_var *) bcom_task_var(tsk->tasknum); inc =3D (struct bcom_gen_bd_tx_inc *) bcom_task_inc(tsk->tasknum);
=20
- if (bcom_load_image(tsk->tasknum, bcom_gen_bd_tx_task))
+ if (bcom_load_image(tsk->tasknum, priv->microcode))
return -1;
=20
var->enable =3D bcom_eng->regs_base +
@@ -253,6 +256,15 @@ bcom_gen_bd_tx_release(struct bcom_task=20
}
EXPORT_SYMBOL_GPL(bcom_gen_bd_tx_release);
=20
+void
+bcom_gen_bd_set_microcode(struct bcom_task *tsk, u32 *microcode) +{
+ struct bcom_gen_bd_priv *priv =3D tsk->priv;
+
+ priv->microcode =3D microcode;
+}
+EXPORT_SYMBOL_GPL(bcom_gen_bd_set_microcode);
+
/* ---------------------------------------------------------------------
* PSC support code
*/
diff -uprN -X linux-2.6.32-rc3/Documentation/dontdiff linux-2.6.32-rc3-orig= /arch/powerpc/sysdev/bestcomm/gen_bd.h linux-2.6.32-rc3/arch/powerpc/sysdev= /bestcomm/gen_bd.h
--- linux-2.6.32-rc3-orig/arch/powerpc/sysdev/bestcomm/gen_bd.h 2009-10-05 =
02:12:30.000000000 +0200
+++ linux-2.6.32-rc3/arch/powerpc/sysdev/bestcomm/gen_bd.h 2009-10-05 14:52=
:31.000000000 +0200
@@ -43,6 +43,9 @@ bcom_gen_bd_tx_reset(struct bcom_task *t
extern void
bcom_gen_bd_tx_release(struct bcom_task *tsk);
=20
+extern void
+bcom_gen_bd_set_microcode(struct bcom_task *tsk, u32 *microcode); +
=20
/* PSC support utility wrappers */
struct bcom_task * bcom_psc_gen_bd_rx_init(unsigned psc_num, int queue_len=
,

fined ones for the MPC52xx's BestComm General Buffer Descriptor (gen_db) ta= sks. =A0The default microcode is still the one from bcom_gen_bd_[rt]x_task,=
but it can be replaced by calling bcom_gen_bd_set_microcode() which is mor=
e efficient than explicitly loading it via bcom_load_image() after each bco= m_gen_bd_[rt]x_reset().

e are no diffierences to the initial attempt.

Hmmm. I've not been comfortable with this change, but it took me a while to put my finger on exactly why. In principle, I think it is a good idea. However, I don't want to merge it without any in-tree users.

The other concern is that I don't like that the patch is only applicable to gen_bd tasks. I've been thinking about this for a while, and I'm not exactly thrilled with the bestcomm layout which keeps the bestcomm firmware separate from the only driver that actually uses it (ie FEC firmware & support code is separate from the FEC driver. Same for ATA). I don't like the fact that code which is only ever used by the ATA driver is maintained completely separate from it. But I may be splitting hairs here and maybe I shouldn't be too concerned about it. Regardless, I'd think I'd rather see something that isn't totally gen_bd specific. Perhaps *microcode should be part of the bcom_task structure?

Thoughts?

g.

ig/arch/powerpc/sysdev/bestcomm/gen_bd.c linux-2.6.32-rc3/arch/powerpc/sysd= ev/bestcomm/gen_bd.c

5 02:12:30.000000000 +0200

9-10-05 14:52:31.000000000 +0200

asknum);

asknum);

asknum);

asknum);

--

ig/arch/powerpc/sysdev/bestcomm/gen_bd.h linux-2.6.32-rc3/arch/powerpc/sysd= ev/bestcomm/gen_bd.h

5 02:12:30.000000000 +0200

9-10-05 14:52:31.000000000 +0200

_len,

--=20
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.

Am 05.10.09 20:16 schrieb(en) Grant Likely:

I could provide my modified microcode (which differs in two words only), for the "standard" task, just with byte swapping for the LPB, but I wonder if it is of much use. No progress with crc32 calculation in BestComm so far... :-(

IMHO the big difference between gen_bd on one hand and ATA and FEC on the other is that (afaik) we have *no* other choice than using the implemented tasks for the latter two, obviously using the Freescale's example microcode. Gen_bd can apparently be used for a plenty of purposes, including LPB (as shown in your driver), maybe PCI (there are examples in Freescale's code), PSC, etc. Thus, the FEC and ATA drivers should not have an option as to change the microcode, whereas it is just convenient for gen_bd.

That's a good point! The separate microcode files for ATA and FEC are somewhat confusing. I guess it's a result of simply copying the Freescale examples into separate files (and that might be a good reason for keeping the microcode separated). Maybe create new sub-folders below arch/powerpc/sysdev/bestcomm for fec, ata, and gen_bd? At least some documentation would be great.

See above - I believe FEC and ATA are somewhat special as they need *exactly* the provided microcode. And if something cannot be changed without breaking the functionality, there should not be a chance to do so... ;-)

Just my €0.01, though...

Cheers, Albrecht.

On Mon, Oct 5, 2009 at 1:42 PM, Albrecht Dre=DF <[email protected]> w= rote:

ile

dea.

for

it

..

d

de.

(as

,

o

agreed.

the

that

ale

No, that doesn't really help much. The ATA and FEC microcode + support functions really have no business living in the common bestcomm directory at all. I should look into moving that stuff right into the FEC and ATA drivers. The Bestcomm API is probably designed to the wrong level, and the base API should expose the same task loading interface to all uses, FEC, ATA, GenBD, and custom tasks.

otally

...

Hence the FEC and ATA microcode should live in (*gasp*) the FEC and ATA drivers. ;-)

g.

--=20
Grant Likely, B.Sc., P.Eng.
Secret Lab Technologies Ltd.