devicetree.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Alexey Kardashevskiy <aik@ozlabs.ru>
To: Gavin Shan <gwshan@linux.vnet.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org, linux-pci@vger.kernel.org,
	devicetree@vger.kernel.org, benh@kernel.crashing.org,
	mpe@ellerman.id.au, dja@axtens.net, bhelgaas@google.com,
	robherring2@gmail.com, grant.likely@linaro.org
Subject: Re: [PATCH v8 08/45] powerpc/powernv: Fix initial IO and M32 segmap
Date: Wed, 13 Apr 2016 19:53:16 +1000	[thread overview]
Message-ID: <570E170C.1040301@ozlabs.ru> (raw)
In-Reply-To: <20160413075324.GA31320@gwshan>

On 04/13/2016 05:53 PM, Gavin Shan wrote:
> On Wed, Apr 13, 2016 at 04:21:07PM +1000, Alexey Kardashevskiy wrote:
>> On 02/17/2016 02:43 PM, Gavin Shan wrote:
>>> There are two arrays for IO and M32 segment maps on every PHB.
>>> The index of the arrays are segment number and the value stored
>>> in the corresponding element is PE number, indicating the segment
>>> is assigned to the PE. Initially, all elements in those two arrays
>>> are zeroes, meaning all segments are assigned to PE#0. It's wrong.
>>>
>>> This fixes the initial values in the elements of those two arrays
>>> to IODA_INVALID_PE, meaning all segments aren't assigned to any
>>> PE.
>>
>> This is ok.
>>
>>> In order to use IODA_INVALID_PE (-1) to represent invalid PE
>>> number, the types of those two arrays are changed from "unsigned int"
>>> to "int".
>>
>> "unsigned" can carry (-1) perfectly fine, just add a type cast to
>> IODA_INVALID_PE:
>>
>> #define IODA_INVALID_PE    (unsigned int)(-1)
>>
>> Using "signed" type for indexes which cannot be negative does not make much
>> sense - instead of checking for the upper boundary, you have to check for "<
>> 0" too.
>>
>> OPAL uses unsigned type for PE (uint64_t or uint32_t or uint16_t - this is
>> quite funny).
>>
>> pnv_ioda_pe::pe_number is "unsigned" and this pe_number is the same thing as
>> I can see in pnv_ioda_setup_dev_PE().
>>
>> Some printk() print the PE number as "%x" (which implies "unsigned").
>>
>
> Yes, I can simply have something like below when PE number as well as
> segment index are represented by "unsigned int" values, right?
>
> #define IODA_INVALID_PE		0xffffffff


This will work too, yes.

>
>>
>> I suggest changing the pci_dn::pe_number type from "int" to "unsigned int" to
>> match pnv_ioda_pe::pe_number, in a separate patch. Or do not touch types for
>> now.
>>
>
> Yes, I will have a separate patch right before this one to address it.
>
>>
>>> Signed-off-by: Gavin Shan <gwshan@linux.vnet.ibm.com>
>>> ---
>>>   arch/powerpc/platforms/powernv/pci-ioda.c | 9 +++++++--
>>>   arch/powerpc/platforms/powernv/pci.h      | 4 ++--
>>>   2 files changed, 9 insertions(+), 4 deletions(-)
>>>
>>> diff --git a/arch/powerpc/platforms/powernv/pci-ioda.c b/arch/powerpc/platforms/powernv/pci-ioda.c
>>> index 1d2514f..44cc5f3 100644
>>> --- a/arch/powerpc/platforms/powernv/pci-ioda.c
>>> +++ b/arch/powerpc/platforms/powernv/pci-ioda.c
>>> @@ -3239,7 +3239,7 @@ static void __init pnv_pci_init_ioda_phb(struct device_node *np,
>>>   	unsigned long size, m32map_off, pemap_off, iomap_off = 0;
>>>   	const __be64 *prop64;
>>>   	const __be32 *prop32;
>>> -	int len;
>>> +	int i, len;
>>>   	u64 phb_id;
>>>   	void *aux;
>>>   	long rc;
>>> @@ -3334,8 +3334,13 @@ static void __init pnv_pci_init_ioda_phb(struct device_node *np,
>>>   	aux = memblock_virt_alloc(size, 0);
>>>   	phb->ioda.pe_alloc = aux;
>>>   	phb->ioda.m32_segmap = aux + m32map_off;
>>> -	if (phb->type == PNV_PHB_IODA1)
>>> +	for (i = 0; i < phb->ioda.total_pe_num; i++)
>>> +		phb->ioda.m32_segmap[i] = IODA_INVALID_PE;
>>> +	if (phb->type == PNV_PHB_IODA1) {
>>>   		phb->ioda.io_segmap = aux + iomap_off;
>>> +		for (i = 0; i < phb->ioda.total_pe_num; i++)
>>> +			phb->ioda.io_segmap[i] = IODA_INVALID_PE;
>>> +	}
>>>   	phb->ioda.pe_array = aux + pemap_off;
>>>   	set_bit(phb->ioda.reserved_pe_idx, phb->ioda.pe_alloc);
>>>
>>> diff --git a/arch/powerpc/platforms/powernv/pci.h b/arch/powerpc/platforms/powernv/pci.h
>>> index 784882a..36c4965 100644
>>> --- a/arch/powerpc/platforms/powernv/pci.h
>>> +++ b/arch/powerpc/platforms/powernv/pci.h
>>> @@ -146,8 +146,8 @@ struct pnv_phb {
>>>   		struct pnv_ioda_pe	*pe_array;
>>>
>>>   		/* M32 & IO segment maps */
>>> -		unsigned int		*m32_segmap;
>>> -		unsigned int		*io_segmap;
>>> +		int			*m32_segmap;
>>> +		int			*io_segmap;
>>>
>>>   		/* IRQ chip */
>>>   		int			irq_chip_init;
>>>
>>
>>
>> --
>> Alexey
>>
>


-- 
Alexey

  reply	other threads:[~2016-04-13  9:53 UTC|newest]

Thread overview: 152+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2016-02-17  3:43 [PATCH v8 00/45] powerpc/powernv: PCI hotplug support Gavin Shan
2016-02-17  3:43 ` [PATCH v8 01/45] PCI: Add pcibios_setup_bridge() Gavin Shan
2016-02-17  3:43 ` [PATCH v8 02/45] powerpc/pci: Override pcibios_setup_bridge() Gavin Shan
     [not found]   ` <1455680668-23298-3-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-13  5:52     ` Alexey Kardashevskiy
2016-02-17  3:43 ` [PATCH v8 03/45] powerpc/pci: Cleanup on struct pci_controller_ops Gavin Shan
2016-02-17  4:18   ` Andrew Donnellan
     [not found]   ` <1455680668-23298-4-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-13  5:52     ` Alexey Kardashevskiy
2016-04-19 23:59       ` Gavin Shan
2016-02-17  3:43 ` [PATCH v8 04/45] powerpc/powernv: Cleanup on pci_controller_ops instances Gavin Shan
2016-02-17  4:38   ` Andrew Donnellan
2016-02-17  3:43 ` [PATCH v8 06/45] powerpc/powernv: Reorder fields in struct pnv_phb Gavin Shan
2016-04-13  5:56   ` Alexey Kardashevskiy
2016-02-17  3:43 ` [PATCH v8 07/45] powerpc/powernv: Rename PE# " Gavin Shan
2016-04-13  5:57   ` Alexey Kardashevskiy
2016-02-17  3:43 ` [PATCH v8 08/45] powerpc/powernv: Fix initial IO and M32 segmap Gavin Shan
2016-04-13  6:21   ` Alexey Kardashevskiy
2016-04-13  7:53     ` Gavin Shan
2016-04-13  9:53       ` Alexey Kardashevskiy [this message]
2016-02-17  3:43 ` [PATCH v8 09/45] powerpc/powernv: Simplify pnv_ioda_setup_pe_seg() Gavin Shan
2016-04-13  6:45   ` Alexey Kardashevskiy
2016-04-20  0:04     ` Gavin Shan
2016-02-17  3:43 ` [PATCH v8 10/45] powerpc/powernv: IO and M32 mapping based on PCI device resources Gavin Shan
2016-02-17  3:43 ` [PATCH v8 11/45] powerpc/powernv: Track M64 segment consumption Gavin Shan
2016-04-13  7:09   ` Alexey Kardashevskiy
2016-04-20  0:05     ` Gavin Shan
2016-02-17  3:43 ` [PATCH v8 12/45] powerpc/powernv: Rename M64 related functions Gavin Shan
     [not found]   ` <1455680668-23298-13-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-13  7:20     ` Alexey Kardashevskiy
2016-02-17  3:43 ` [PATCH v8 14/45] powerpc/powernv/ioda1: Rename pnv_pci_ioda_setup_dma_pe() Gavin Shan
2016-04-13  7:36   ` Alexey Kardashevskiy
2016-02-17  3:43 ` [PATCH v8 16/45] powerpc/powernv: Remove DMA32 PE list Gavin Shan
2016-04-13  8:59   ` Alexey Kardashevskiy
2016-04-20  0:34     ` Gavin Shan
     [not found] ` <1455680668-23298-1-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-02-17  3:43   ` [PATCH v8 05/45] powerpc/powernv: Drop phb->bdfn_to_pe() Gavin Shan
2016-04-13  5:53     ` Alexey Kardashevskiy
2016-02-17  3:43   ` [PATCH v8 13/45] powerpc/powernv/ioda1: M64 support on P7IOC Gavin Shan
2016-04-13  7:47     ` Alexey Kardashevskiy
2016-04-20  0:22       ` Gavin Shan
2016-04-20  2:55         ` Alexey Kardashevskiy
2016-02-17  3:43   ` [PATCH v8 15/45] powerpc/powernv/ioda1: Introduce PNV_IODA1_DMA32_SEGSIZE Gavin Shan
2016-04-13  8:29     ` Alexey Kardashevskiy
2016-04-13 23:54       ` Gavin Shan
2016-04-14  3:36         ` Alexey Kardashevskiy
2016-04-20  0:25           ` Gavin Shan
2016-02-17  3:44   ` [PATCH v8 17/45] powerpc/powernv/ioda1: Improve DMA32 segment track Gavin Shan
2016-04-19  1:50     ` Alexey Kardashevskiy
2016-04-20  0:49       ` Gavin Shan
2016-04-20  5:10         ` Alexey Kardashevskiy
2016-02-17  3:44   ` [PATCH v8 38/45] powerpc/powernv: Functions to get/set PCI slot status Gavin Shan
2016-04-19  9:39     ` Alexey Kardashevskiy
2016-04-20  2:36       ` Gavin Shan
2016-04-20  4:25         ` Alexey Kardashevskiy
2016-02-17  3:44   ` [PATCH v8 42/45] drivers/of: Rename unflatten_dt_node() Gavin Shan
     [not found]     ` <1455680668-23298-43-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-02-17 14:59       ` Rob Herring
2016-02-19  3:16         ` Gavin Shan
2016-03-02  2:40           ` Rob Herring
2016-03-08  0:56             ` Gavin Shan
2016-03-17 13:31               ` Rob Herring
2016-03-17 22:44                 ` Gavin Shan
2016-02-17  3:44   ` [PATCH v8 45/45] PCI/hotplug: PowerPC PowerNV PCI hotplug driver Gavin Shan
2016-04-15  0:47     ` Alistair Popple
2016-04-15  1:39       ` Gavin Shan
     [not found]     ` <1455680668-23298-46-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-19 10:36       ` Alexey Kardashevskiy
2016-04-20  1:55         ` Alistair Popple
2016-05-02 23:41           ` Gavin Shan
2016-05-03  0:44             ` Alexey Kardashevskiy
2016-05-02  3:44         ` Gavin Shan
2016-05-02  6:11           ` Alexey Kardashevskiy
2016-05-02 23:38             ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 18/45] powerpc/powernv: Increase PE# capacity Gavin Shan
2016-04-19  2:02   ` Alexey Kardashevskiy
2016-04-20  0:52     ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 19/45] powerpc/powernv: Use PE instead of number during setup and release Gavin Shan
2016-04-19  2:50   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 20/45] powerpc/powernv: Allocate PE# in reverse order Gavin Shan
2016-04-19  3:07   ` Alexey Kardashevskiy
2016-04-20  1:04     ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 21/45] powerpc/powernv: Create PEs at PCI hot plugging time Gavin Shan
2016-04-19  4:16   ` Alexey Kardashevskiy
2016-04-20  1:12     ` Gavin Shan
2016-04-20  3:00       ` Alexey Kardashevskiy
2016-04-20  3:35         ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 22/45] powerpc/powernv/ioda1: Support releasing IODA1 TCE table Gavin Shan
2016-04-19  4:28   ` Alexey Kardashevskiy
2016-04-20  1:15     ` Gavin Shan
2016-04-20  3:17       ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 23/45] powerpc/powernv: Dynamically release PEs Gavin Shan
2016-04-19  5:19   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 24/45] powerpc/pci: Rename pcibios_{add,remove}_pci_devices() Gavin Shan
2016-04-19  5:28   ` Alexey Kardashevskiy
2016-04-20  1:23     ` Gavin Shan
2016-04-20  3:21       ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 25/45] powerpc/pci: Rename pcibios_find_pci_bus() Gavin Shan
2016-04-19  5:31   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 26/45] powerpc/pci: Move pci_find_bus_by_node() around Gavin Shan
2016-02-17  3:44 ` [PATCH v8 27/45] powerpc/pci: Export pci_add_device_node_info() Gavin Shan
2016-04-19  5:35   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 28/45] powerpc/pci: Introduce pci_remove_device_node_info() Gavin Shan
2016-04-19  5:48   ` Alexey Kardashevskiy
2016-04-20  1:25     ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 29/45] powerpc/pci: Export pci_traverse_device_nodes() Gavin Shan
     [not found]   ` <1455680668-23298-30-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-19  5:51     ` Alexey Kardashevskiy
2016-04-20  1:27       ` Gavin Shan
2016-04-20  3:39         ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 30/45] powerpc/pci: Delay populating pdn Gavin Shan
2016-04-19  8:19   ` Alexey Kardashevskiy
2016-04-20  2:13     ` Gavin Shan
2016-04-20  3:54       ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 31/45] powerpc/pci: Don't scan empty slot Gavin Shan
2016-04-19  8:19   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 32/45] powerpc/pci: Update bridge windows on PCI plug Gavin Shan
2016-04-19  8:47   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 33/45] powerpc/powernv: Simplify pnv_eeh_reset() Gavin Shan
2016-02-17  4:35   ` Andrew Donnellan
2016-04-19  8:49   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 34/45] powerpc/powernv: Exclude root bus in pnv_pci_reset_secondary_bus() Gavin Shan
2016-04-19  8:57   ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 35/45] powerpc/powernv: Fundamental reset " Gavin Shan
     [not found]   ` <1455680668-23298-36-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-19  9:04     ` Alexey Kardashevskiy
2016-04-20  1:36       ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 36/45] powerpc/powernv: Support PCI slot ID Gavin Shan
     [not found]   ` <1455680668-23298-37-git-send-email-gwshan-23VcF4HTsmIX0ybBhKVfKdBPR1lH4CV8@public.gmane.org>
2016-04-19  9:28     ` Alexey Kardashevskiy
2016-04-20  2:28       ` Gavin Shan
2016-04-20  4:14         ` Alexey Kardashevskiy
2016-04-22  4:23           ` Alistair Popple
2016-02-17  3:44 ` [PATCH v8 37/45] powerpc/powernv: Use firmware PCI slot reset infrastructure Gavin Shan
2016-04-19  9:34   ` Alexey Kardashevskiy
2016-04-20  2:33     ` Gavin Shan
2016-04-20  4:17       ` Alexey Kardashevskiy
2016-02-17  3:44 ` [PATCH v8 39/45] powerpc/powernv: Select OF_DYNAMIC Gavin Shan
2016-04-19  9:42   ` Alexey Kardashevskiy
2016-04-20  2:38     ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 40/45] drivers/of: Split unflatten_dt_node() Gavin Shan
2016-02-17 14:30   ` Rob Herring
2016-04-20  2:38     ` Gavin Shan
2016-05-02  2:02     ` Gavin Shan
2016-02-17  3:44 ` [PATCH v8 41/45] drivers/of: Avoid recursively calling unflatten_dt_node() Gavin Shan
2016-02-17 14:53   ` Rob Herring
2016-02-17  3:44 ` [PATCH v8 43/45] drivers/of: Specify parent node in of_fdt_unflatten_tree() Gavin Shan
2016-02-17 15:00   ` Rob Herring
2016-02-17 15:58   ` Jyri Sarha
2016-02-17  3:44 ` [PATCH v8 44/45] drivers/of: Return allocated memory from of_fdt_unflatten_tree() Gavin Shan
2016-04-13  7:28 ` [PATCH v8 00/45] powerpc/powernv: PCI hotplug support Alexey Kardashevskiy
2016-04-13  7:42   ` Gavin Shan
2016-04-13  9:14     ` Alexey Kardashevskiy
2016-04-13 23:42       ` Gavin Shan
2016-04-13 23:57         ` Alistair Popple
2016-04-14  1:30           ` Gavin Shan
2016-04-14  3:38             ` Alexey Kardashevskiy
2016-04-15 16:10             ` Rob Herring
2016-04-20  2:40               ` Gavin Shan
2016-04-14  3:26         ` Alexey Kardashevskiy
2016-04-14  5:25           ` Gavin Shan

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=570E170C.1040301@ozlabs.ru \
    --to=aik@ozlabs.ru \
    --cc=benh@kernel.crashing.org \
    --cc=bhelgaas@google.com \
    --cc=devicetree@vger.kernel.org \
    --cc=dja@axtens.net \
    --cc=grant.likely@linaro.org \
    --cc=gwshan@linux.vnet.ibm.com \
    --cc=linux-pci@vger.kernel.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=mpe@ellerman.id.au \
    --cc=robherring2@gmail.com \
    /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 a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).