From: Anton Vorontsov <avorontsov@ru.mvista.com>
To: Kumar Gala <galak@kernel.crashing.org>
Cc: Paul Mackerras <paulus@samba.org>, linuxppc-dev@ozlabs.org
Subject: [PATCH v2 3/4] powerpc/fsl_pci: Add MPC83xx PCI-E controller RC mode support
Date: Tue, 30 Dec 2008 00:05:59 +0300 [thread overview]
Message-ID: <20081229210559.GA10801@oksana.dev.rtsoft.ru> (raw)
In-Reply-To: <DE5232AE-281E-4A6E-9844-A69DF7C90034@kernel.crashing.org>
This patch adds support for PCI-Express controllers as found on the
newer MPC83xx chips.
The work is loosely based on the Tony Li's patch[1], but unlike the
original patch, this patch implements sliding window for the Type 1
transactions using outbound window translations, so we don't have to
ioremap the whole PCI-E configuration space.
[1] http://ozlabs.org/pipermail/linuxppc-dev/2008-January/049028.html
Signed-off-by: Tony Li <tony.li@freescale.com>
Signed-off-by: Anton Vorontsov <avorontsov@ru.mvista.com>
---
On Mon, Dec 29, 2008 at 12:09:38PM -0600, Kumar Gala wrote:
>
> On Dec 29, 2008, at 10:40 AM, Anton Vorontsov wrote:
>
>> This patch adds support for PCI-Express controllers as found on the
>> newer MPC83xx chips.
>>
>> The work is losely based on the Tony Li's patch[1], but unlike the
>> original patch, this patch implements sliding window for the Type 1
>> configuration transactions, so we don't have to ioremap the whole
>> PCI-E configuration space.
>>
>> Since we have to remap the config space in the atomic context, and
>> as ioremap() function might sleep, we use the fixmap API. With fixmap
>> we can map the page-sized areas in the atomic context. The approch
>> is similar to the x86 implementation (arch/x86/pci/mmconfig_32.c).
>>
>> [1] http://ozlabs.org/pipermail/linuxppc-dev/2008-January/049028.html
>
> thanks for looking at cleaning this up. Have you looked at moving the
> outbound window as well? Its not a critical issue but something I was
> thinking about.
Ah, I think we can indeed slide the outbound window instead. That
way we don't need the fixmap at all.
See how I change the PEX_OUTWIN0_TAL register in the new
mpc83xx_pcie_remap_cfg:
According to specs, bits 7:2 are used to offset the register,
bits 14:12 are used to denote a function, 19:15 are used for
device number, 20:27 should specify bus number (original code
was writing the bus number to the wrong bit field, I think).
So, we ioremap the first 4kb of the config space at the setup time,
and then just changing the TAL to read/write config space of the
needed BDF.
[...]
>> +#define PCI_DEVICE_ID_MPC8377 0x00c7
>
> Please make sure these get added to http://pci-ids.ucw.cz/
Will do.
How about this patch? It works for me using SKY2 NIC, though I don't
have any PCI-E switch or multi-function device to fully test the
sliding window behaviour...
Thanks,
arch/powerpc/include/asm/pci-bridge.h | 1 +
arch/powerpc/sysdev/fsl_pci.c | 195 ++++++++++++++++++++++++++++++++-
include/linux/pci_ids.h | 8 ++
3 files changed, 200 insertions(+), 4 deletions(-)
diff --git a/arch/powerpc/include/asm/pci-bridge.h b/arch/powerpc/include/asm/pci-bridge.h
index 84007af..b4a9e68 100644
--- a/arch/powerpc/include/asm/pci-bridge.h
+++ b/arch/powerpc/include/asm/pci-bridge.h
@@ -132,6 +132,7 @@ struct pci_controller {
#define PPC_INDIRECT_TYPE_NO_PCIE_LINK 0x00000008
#define PPC_INDIRECT_TYPE_BIG_ENDIAN 0x00000010
#define PPC_INDIRECT_TYPE_BROKEN_MRM 0x00000020
+#define PPC_INDIRECT_TYPE_MPC83XX_PCIE 0x00000040
u32 indirect_type;
#endif /* !CONFIG_PPC64 */
/* Currently, we limit ourselves to 1 IO range and 3 mem
diff --git a/arch/powerpc/sysdev/fsl_pci.c b/arch/powerpc/sysdev/fsl_pci.c
index f611d03..9e104f8 100644
--- a/arch/powerpc/sysdev/fsl_pci.c
+++ b/arch/powerpc/sysdev/fsl_pci.c
@@ -1,12 +1,16 @@
/*
* MPC83xx/85xx/86xx PCI/PCIE support routing.
*
- * Copyright 2007,2008 Freescale Semiconductor, Inc
+ * Copyright 2007-2009 Freescale Semiconductor, Inc.
+ * Copyright 2008-2009 MontaVista Software, Inc.
*
* Initial author: Xianghua Xiao <x.xiao@freescale.com>
* Recode: ZHANG WEI <wei.zhang@freescale.com>
* Rewrite the routing for Frescale PCI and PCI Express
* Roy Zang <tie-fei.zang@freescale.com>
+ * MPC83xx PCI-Express support:
+ * Tony Li <tony.li@freescale.com>
+ * Anton Vorontsov <avorontsov@ru.mvista.com>
*
* This program is free software; you can redistribute it and/or modify it
* under the terms of the GNU General Public License as published by the
@@ -27,7 +31,6 @@
#include <sysdev/fsl_soc.h>
#include <sysdev/fsl_pci.h>
-#if defined(CONFIG_PPC_85xx) || defined(CONFIG_PPC_86xx)
/* atmu setup for fsl pci/pcie controller */
static void __init setup_pci_atmu(struct pci_controller *hose,
struct resource *rsrc)
@@ -228,6 +231,14 @@ int __init fsl_add_bridge(struct device_node *dev, int is_primary)
return 0;
}
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8314E, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8314, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8315E, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8315, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8377E, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8377, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8378E, quirk_fsl_pcie_header);
+DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8378, quirk_fsl_pcie_header);
DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8548E, quirk_fsl_pcie_header);
DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8548, quirk_fsl_pcie_header);
DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8543E, quirk_fsl_pcie_header);
@@ -250,9 +261,169 @@ DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8536, quirk_fsl_pcie_header);
DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8641, quirk_fsl_pcie_header);
DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8641D, quirk_fsl_pcie_header);
DECLARE_PCI_FIXUP_HEADER(0x1957, PCI_DEVICE_ID_MPC8610, quirk_fsl_pcie_header);
-#endif /* CONFIG_PPC_85xx || CONFIG_PPC_86xx */
#if defined(CONFIG_PPC_83xx) || defined(CONFIG_PPC_MPC512x)
+struct mpc83xx_pcie {
+ void __iomem *cfg_map;
+ u32 dev_base;
+};
+
+/*
+ * With the convention of u-boot, the PCIE outbound window 0 serves
+ * as configuration transactions outbound.
+ */
+#define PEX_OUTWIN0_TAL 0xCA8
+#define PEX_OUTWIN0_TAH 0xCAC
+
+static int mpc83xx_pcie_exclude_device(struct pci_bus *bus, unsigned int devfn)
+{
+ struct pci_controller *hose = bus->sysdata;
+
+ if (hose->indirect_type & PPC_INDIRECT_TYPE_NO_PCIE_LINK)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+ /*
+ * Workaround for the HW bug: for Type 0 configure transactions the
+ * PCI-E controller does not check the device number bits and just
+ * assumes that the device number bits are 0.
+ */
+ if (hose->indirect_type & PPC_INDIRECT_TYPE_MPC83XX_PCIE) {
+ if (bus->number == hose->first_busno ||
+ bus->primary == hose->first_busno) {
+ if (devfn & 0xf8)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+ }
+ }
+
+ if (ppc_md.pci_exclude_device) {
+ if (ppc_md.pci_exclude_device(hose, bus->number, devfn))
+ return PCIBIOS_DEVICE_NOT_FOUND;
+ }
+
+ return PCIBIOS_SUCCESSFUL;
+}
+
+static void __iomem *mpc83xx_pcie_remap_cfg(struct pci_bus *bus,
+ unsigned int devfn, int offset)
+{
+ struct pci_controller *hose = bus->sysdata;
+ struct mpc83xx_pcie *pcie = hose->dn->data;
+ u8 bus_no = bus->number - hose->first_busno;
+ u32 dev_base = bus_no << 20 | devfn << 12;
+ int ret;
+
+ ret = mpc83xx_pcie_exclude_device(bus, devfn);
+ if (ret)
+ return NULL;
+
+ offset &= 0xfff;
+
+ if (hose->indirect_type & PPC_INDIRECT_TYPE_MPC83XX_PCIE) {
+ /* Type 0 */
+ if (bus->number == hose->first_busno)
+ return hose->cfg_data + offset;
+ }
+
+ if (pcie->dev_base == dev_base)
+ goto mapped;
+
+ out_le32(hose->cfg_data + PEX_OUTWIN0_TAL, dev_base);
+
+ pcie->dev_base = dev_base;
+mapped:
+ return pcie->cfg_map + offset;
+}
+
+static int mpc83xx_pcie_read_config(struct pci_bus *bus, unsigned int devfn,
+ int offset, int len, u32 *val)
+{
+ void __iomem *cfg_addr;
+
+ cfg_addr = mpc83xx_pcie_remap_cfg(bus, devfn, offset);
+ if (!cfg_addr)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+
+ switch (len) {
+ case 1:
+ *val = in_8(cfg_addr);
+ break;
+ case 2:
+ *val = in_le16(cfg_addr);
+ break;
+ default:
+ *val = in_le32(cfg_addr);
+ break;
+ }
+
+ return PCIBIOS_SUCCESSFUL;
+}
+
+static int mpc83xx_pcie_write_config(struct pci_bus *bus, unsigned int devfn,
+ int offset, int len, u32 val)
+{
+ void __iomem *cfg_addr;
+
+ cfg_addr = mpc83xx_pcie_remap_cfg(bus, devfn, offset);
+ if (!cfg_addr)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+
+ switch (len) {
+ case 1:
+ out_8(cfg_addr, val);
+ break;
+ case 2:
+ out_le16(cfg_addr, val);
+ break;
+ default:
+ out_le32(cfg_addr, val);
+ break;
+ }
+
+ return PCIBIOS_SUCCESSFUL;
+}
+
+static struct pci_ops mpc83xx_pcie_ops = {
+ mpc83xx_pcie_read_config,
+ mpc83xx_pcie_write_config,
+};
+
+static int __init mpc83xx_pcie_setup(struct pci_controller *hose,
+ struct resource *reg,
+ struct resource *cfg)
+{
+ struct mpc83xx_pcie *pcie;
+
+ pcie = zalloc_maybe_bootmem(sizeof(*pcie), GFP_KERNEL);
+ if (!pcie)
+ return -ENOMEM;
+
+ pcie->cfg_map = ioremap(cfg->start, 0x1000);
+ if (!pcie->cfg_map)
+ goto err_cfg_map;
+
+ hose->cfg_data = ioremap(reg->start, resource_size(reg));
+ if (!hose->cfg_data)
+ goto err_cfg_data;
+
+ WARN_ON(hose->dn->data);
+ hose->dn->data = pcie;
+ hose->indirect_type |= PPC_INDIRECT_TYPE_MPC83XX_PCIE;
+ hose->ops = &mpc83xx_pcie_ops;
+
+ out_le32(hose->cfg_data + PEX_OUTWIN0_TAH, 0);
+ out_le32(hose->cfg_data + PEX_OUTWIN0_TAL, 0);
+
+ if (fsl_pcie_check_link(hose))
+ hose->indirect_type |= PPC_INDIRECT_TYPE_NO_PCIE_LINK;
+
+ return 0;
+err_cfg_data:
+ iounmap(pcie->cfg_map);
+err_cfg_map:
+ kfree(pcie);
+ return -ENOMEM;
+
+}
+
int __init mpc83xx_add_bridge(struct device_node *dev)
{
int len;
@@ -262,6 +433,10 @@ int __init mpc83xx_add_bridge(struct device_node *dev)
const int *bus_range;
int primary;
+ if (!of_device_is_available(dev)) {
+ pr_warning("%s: disabled by the firmware.\n", dev->full_name);
+ return -ENODEV;
+ }
pr_debug("Adding PCI host bridge %s\n", dev->full_name);
/* Fetch host bridge registers address */
@@ -309,7 +484,16 @@ int __init mpc83xx_add_bridge(struct device_node *dev)
hose->first_busno = bus_range ? bus_range[0] : 0;
hose->last_busno = bus_range ? bus_range[1] : 0xff;
- setup_indirect_pci(hose, rsrc_cfg.start, rsrc_cfg.start + 4, 0);
+ if (of_device_is_compatible(dev, "fsl,mpc8314-pcie")) {
+ int ret;
+
+ ret = mpc83xx_pcie_setup(hose, &rsrc_reg, &rsrc_cfg);
+ if (ret)
+ goto err;
+ } else {
+ setup_indirect_pci(hose, rsrc_cfg.start,
+ rsrc_cfg.start + 4, 0);
+ }
printk(KERN_INFO "Found FSL PCI host bridge at 0x%016llx. "
"Firmware bus number: %d->%d\n",
@@ -324,5 +508,8 @@ int __init mpc83xx_add_bridge(struct device_node *dev)
pci_process_bridge_OF_ranges(hose, dev, primary);
return 0;
+err:
+ pcibios_free_controller(hose);
+ return -ENOMEM;
}
#endif /* CONFIG_PPC_83xx */
diff --git a/include/linux/pci_ids.h b/include/linux/pci_ids.h
index 1800f1d..b30e28f 100644
--- a/include/linux/pci_ids.h
+++ b/include/linux/pci_ids.h
@@ -2209,6 +2209,14 @@
#define PCI_DEVICE_ID_TDI_EHCI 0x0101
#define PCI_VENDOR_ID_FREESCALE 0x1957
+#define PCI_DEVICE_ID_MPC8315E 0x00b4
+#define PCI_DEVICE_ID_MPC8315 0x00b5
+#define PCI_DEVICE_ID_MPC8314E 0x00b6
+#define PCI_DEVICE_ID_MPC8314 0x00b7
+#define PCI_DEVICE_ID_MPC8378E 0x00c4
+#define PCI_DEVICE_ID_MPC8378 0x00c5
+#define PCI_DEVICE_ID_MPC8377E 0x00c6
+#define PCI_DEVICE_ID_MPC8377 0x00c7
#define PCI_DEVICE_ID_MPC8548E 0x0012
#define PCI_DEVICE_ID_MPC8548 0x0013
#define PCI_DEVICE_ID_MPC8543E 0x0014
--
1.5.6.5
next prev parent reply other threads:[~2008-12-29 21:06 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-12-29 16:40 [PATCH 3/4] powerpc/fsl_pci: Add MPC83xx PCI-E controller RC mode support Anton Vorontsov
2008-12-29 18:09 ` Kumar Gala
2008-12-29 21:05 ` Anton Vorontsov [this message]
2008-12-30 18:07 ` [PATCH v2 " Kumar Gala
2008-12-30 20:18 ` [PATCH v3 " Anton Vorontsov
2009-01-05 8:18 ` [PATCH v2 3/4] powerpc/fsl_pci: Add MPC83xx PCI-E controller RCmode support Liu Dave
2009-01-05 8:41 ` Arun Kumar
2009-01-05 16:30 ` Anton Vorontsov
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20081229210559.GA10801@oksana.dev.rtsoft.ru \
--to=avorontsov@ru.mvista.com \
--cc=galak@kernel.crashing.org \
--cc=linuxppc-dev@ozlabs.org \
--cc=paulus@samba.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.