h-4nd-h
/
plx
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
32 Commits
3 Branches
0 Tags
2.9 MiB
 
 
 
Tree: 0a43a11850
Branches Tags
${ item.name }
Create tag ${ searchTerm }
Create branch ${ searchTerm }
from '0a43a11850'
${ noResults }
plx/Plxlibrary/Plx_sysdep.h

490 lines
17 KiB
Raw Blame History

#ifndef _PLX_SYSDEP_H_
#define _PLX_SYSDEP_H_
/*******************************************************************************
* Copyright 2013-2019 Broadcom Inc
* Copyright (c) 2009 to 2012 PLX Technology Inc. All rights reserved.
*
* This software is available to you under a choice of one of two
* licenses. You may choose to be licensed under the terms of the GNU
* General Public License (GPL) Version 2, available from the file
* COPYING in the main directorY of this source tree, or the
* BSD license below:
*
* Redistribution and use in source and binary forms, with or
* without modification, are permitted provided that the following
* conditions are met:
*
* - Redistributions of source code must retain the above
* copyright notice, this list of conditions and the following
* disclaimer.
*
* - Redistributions in binary form must reproduce the above
* copyright notice, this list of conditions and the following
* disclaimer in the documentation and/or other materials
* provided with the distribution.
*
* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND,
* EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
* MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
* NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS
* BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN
* ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN
* CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
* SOFTWARE.
******************************************************************************/
/******************************************************************************
*
* File Name:
*
* Plx_sysdep.h
*
* Description:
*
* This file is provided to support compatible code between different
* Linux kernel versions.
*
* Revision History:
*
* 04-01-19 : PCI/PCIe SDK v8.00
*
*****************************************************************************/
#ifndef LINUX_VERSION_CODE
#include <linux/version.h>
#endif
// Only allow 2.6.18 and higher kernels
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,18)
#error "ERROR: Linux kernel versions prior to v2.6.18 not supported"
#endif
/***********************************************************
* IORESOURCE_MEM_64
*
* This flag specifies whether a PCI BAR space is 64-bit. The
* definition wasn't added until 2.6.31.
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,31))
#define IORESOURCE_MEM_64 0x00100000
#endif
/***********************************************************
* VM_RESERVED
*
* This flag was removed starting with 3.7. The recommended
* replacement is a combination of two flags.
**********************************************************/
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(3,7,0))
#define VM_RESERVED (VM_DONTEXPAND | VM_DONTDUMP)
#endif
/***********************************************************
* INIT_WORK & INIT_DELAYED_WORK
*
* This macro initializes a work structure with the function
* to call. In kernel 2.6.20, the 3rd parameter was removed.
* It used to be the parameter to the function, but now, the
* function is called with a pointer to the work_struct itself.
* INIT_DELAYED_WORK was introduced to init the delayed_work struct.
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20))
#define PLX_INIT_WORK INIT_WORK
// Red Hat pre-defines this
#if !defined(RED_HAT_LINUX_KERNEL)
#define INIT_DELAYED_WORK INIT_WORK
#endif
#else
#define PLX_INIT_WORK(work, func, data) INIT_WORK((work), (func))
#endif
/***********************************************************
* ioremap_prot
*
* This function is supported after 2.6.27 only one some
* architectures, like x86 & PowerPC. Other architectures
* added support for it in later kernels. For platforms
* that support it, HAVE_IOREMAP_PROT is expected defined.
* SDK drivers use the function for probing ACPI tables. In
* newer kernels, calls to ioremap() for ACPI locations may
* report errors if default flags conflict with kernel mappings.
**********************************************************/
#if !defined(CONFIG_HAVE_IOREMAP_PROT)
// Revert to ioremap() for unsupported architectures
#define ioremap_prot(addr,size,flags) ioremap((addr), (size))
#endif
/***********************************************************
* pgprot_writecombine
*
* This function is not supported on all platforms. There is
* a standard definition for it starting with 2.6.29.
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,29))
#define pgprot_writecombine pgprot_noncached
#endif
/***********************************************************
* access_ok
*
* access_ok() removed type param in 5.0
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(5,0,0))
#define Plx_access_ok access_ok
#else
#define Plx_access_ok(type,addr,size) access_ok( (addr),(size) )
#endif
/***********************************************************
* sema_init / init_MUTEX
*
* init_MUTEX replaced by sema_init starting with 2.6.26
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,26))
#define Plx_sema_init(sem, n) init_MUTEX( (sem) )
#else
#define Plx_sema_init sema_init
#endif
/***********************************************************
* flush_work
*
* Flush work queue function not added until 2.6.27
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,27))
#define Plx_flush_work(x)
#else
#define Plx_flush_work flush_work
#endif
/***********************************************************
* PLX_DPC_PARAM
*
* In kernel 2.6.20, the parameter to a work queue function
* was made to always be a pointer to the work_struct itself.
* In previous kernels, this was always a VOID*. Since
* PLX drivers use work queue functions for the DPC/bottom-half
* processing, the parameter had to be changed. For cleaner
* source code, the definition PLX_DPC_PARAM is used and is
* defined below.
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,20))
#define PLX_DPC_PARAM VOID
#else
#define PLX_DPC_PARAM struct work_struct
#endif
/***********************************************************
* pci_get_domain_bus_and_slot not added until 2.6.33
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,33))
#define Plx_pci_get_domain_bus_and_slot(d,b,df) pci_get_bus_and_slot( b, df )
#else
#define Plx_pci_get_domain_bus_and_slot pci_get_domain_bus_and_slot
#endif
/***********************************************************
* pci_enable_msi/pci_enable_msix deprecated
*
* The pci_*_msi/msix MSI functions are deprecated as of
* kernel 4.8. A new set of PCI subsystem functions have
* replaced them.
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(4,8,0))
#define Plx_pci_enable_msi pci_enable_msi
#define Plx_pci_disable_msi pci_disable_msi
#define Plx_pci_enable_msix pci_enable_msix
#define Plx_pci_disable_msix pci_disable_msix
#else
#define Plx_pci_enable_msi( pdev ) pci_alloc_irq_vectors( (pdev), 1, 1, PCI_IRQ_MSI )
#define Plx_pci_disable_msi pci_free_irq_vectors
#define Plx_pci_disable_msix pci_free_irq_vectors
#define Plx_pci_enable_msix(pdev,entries,nvec) \
({ \
int _rc; \
int _idx; \
\
/* Attempt to allocate MSI-X vectors */ \
_rc = pci_alloc_irq_vectors( \
(pdev), (nvec), (nvec), PCI_IRQ_MSIX ); \
if (_rc == (nvec)) \
{ \
/* Set successful return value */ \
_rc = 0; \
\
/* Fill in the vector table */ \
for (_idx = 0; _idx < (nvec); _idx++) \
{ \
(entries)[_idx].vector = \
pci_irq_vector( (pdev), (entries)[_idx].entry ); \
} \
} \
\
_rc; \
})
#endif
/***********************************************************
* kmap_atomic/kunmap_atomic - 2nd parameter removed in 2.6.37
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,37))
#define Plx_kmap_atomic kmap_atomic
#define Plx_kunmap_atomic kunmap_atomic
#else
#define Plx_kmap_atomic(page,kmtype) kmap_atomic( (page) )
#define Plx_kunmap_atomic(page,kmtype) kunmap_atomic( (page) )
#endif
/***********************************************************
* 'secondary' & 'subordinate' fields removed from 'struct pci_bus'
* in 3.6 & replaced with start/end fields in a 'struct resource'.
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(3,6,0))
#define Plx_STRUCT_PCI_BUS_SEC(pbus) (pbus)->secondary
#define Plx_STRUCT_PCI_BUS_SUB(pbus) (pbus)->subordinate
#else
#define Plx_STRUCT_PCI_BUS_SEC(pbus) (pbus)->busn_res.start
#define Plx_STRUCT_PCI_BUS_SUB(pbus) (pbus)->busn_res.end
#endif
/***********************************************************
* skb_frag_struct "page" field changed from pointer to
* a page pointer within a structure in 3.2.0
**********************************************************/
#if (LINUX_VERSION_CODE < KERNEL_VERSION(3,2,0))
#define Plx_SKB_FRAG_STRUCT_PAGE(frag) (frag)->page
#else
#define Plx_SKB_FRAG_STRUCT_PAGE(frag) (frag)->page.p
#endif
/***********************************************************
* pcie_cap field in pci_dev - Added by 2.6.38
*
* The pci_dev structure has a pcie_cap field to report the
* device's PCIe capability (10h) starting offset.
* This field is was also added to updated RedHat kernels.
**********************************************************/
#if (LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,38)) || \
((LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,32)) && defined(RED_HAT_LINUX_KERNEL))
#define Plx_pcie_cap(pDev) ((pDev)->pcie_cap)
#else
#define Plx_pcie_cap(pDev) pci_find_capability( (pDev), 0x10 )
#endif
/***********************************************************
* is_virtfn field in pci_dev - Added in 2.6.30
* pci_physfn - Added by 2.6.35
*
* For SR-IOV devices, there is an 'is_virtfn' field in the
* pci_dev structure to report whether the device is a VF. This
* field is was also added to updated RedHat kernels.
*
* pci_physfn returns the parent PF of a VF; otherwise, returns
* the passed device.
***********************************************************/
#if defined(CONFIG_PCI_IOV) && \
((LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,30)) || \
((LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,18)) && defined(RED_HAT_LINUX_KERNEL)))
#define Plx_pcie_is_virtfn(pDev) ((pDev)->is_virtfn)
#if (LINUX_VERSION_CODE < KERNEL_VERSION(2,6,35))
#define Plx_pci_physfn(pDev) \
({ \
struct pci_dev *_pDev; \
\
if (pDev->is_virtfn) \
_pDev = (pDev->physfn); \
else \
_pDev = pDev; \
_pDev; \
})
#else
#define Plx_pci_physfn pci_physfn
#endif
#else
#define Plx_pcie_is_virtfn(pDev) (FALSE)
#define Plx_pci_physfn(pDev) (pDev)
#endif
/***********************************************************
* readq / writeq
*
* These functions are used to perform 64-bit accesses to
* I/O memory. They are not defined for all architectures.
* For x86 32-bit, they were added in 2.6.29.
*
* NOTE: This is still incomplete for non-x86 32-bit architectures
* where readq/writeq are not yet defined.
**********************************************************/
#if ((LINUX_VERSION_CODE < KERNEL_VERSION(2,6,29)) && defined(CONFIG_X86_32))
// x86 64-bit I/O access functions
static inline __u64 readq(const volatile void __iomem *addr)
{
const volatile u32 __iomem *p = addr;
u64 value;
value = readl(p);
value += ((u64)readl(p + 1) << 32);
return value;
}
static inline void writeq(__u64 val, volatile void __iomem *addr)
{
writel(val, addr);
writel(val >> 32, addr+4);
}
#elif !defined(CONFIG_64BIT) && !defined(CONFIG_X86_32)
// Revert to 32-bit accesses for non-x86/x64 platforms
#define readq readl
#define writeq writel
#endif
/***********************************************************
* get_user_pages
*
* Parameters to this function changed as follows:
* 4.6: Removed first two params (curr task & task's mem-manage struct)
* 4.9: Replaced write & force params with a single gup_flags param
**********************************************************/
#if LINUX_VERSION_CODE < KERNEL_VERSION(4,6,0)
#define Plx_get_user_pages(start, nr_pages, gup_flags, pages, vmas) \
( \
get_user_pages( \
current, \
current->mm, \
(start), \
(nr_pages), \
((gup_flags) & FOLL_WRITE) ? 1 : 0, \
0, \
(pages), \
(vmas) \
) \
)
#elif LINUX_VERSION_CODE < KERNEL_VERSION(4,9,0)
#define Plx_get_user_pages(start, nr_pages, gup_flags, pages, vmas) \
( \
get_user_pages( \
(start), \
(nr_pages), \
((gup_flags) & FOLL_WRITE) ? 1 : 0, \
0, \
(pages), \
(vmas) \
) \
)
#else
#define Plx_get_user_pages get_user_pages
#endif
/***********************************************************
* dma_set_coherent_mask / pci_set_consistent_dma_mask
*
* This function is used to set the mask for coherent/consistent
* DMA buffer allocations. Prior to 2.6.34, the function is
* pci_set_consistent_dma_mask. It is now dma_set_coherent_mask.
* The first parameter has also been changed from pci_dev
* structure to the dev structure found in pci_dev.
**********************************************************/
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,34)
#define Plx_dma_set_coherent_mask(pdx, mask) \
( \
pci_set_consistent_dma_mask( \
(pdx)->pPciDevice, \
(mask) \
) \
)
#else
#define Plx_dma_set_coherent_mask(pdx, mask) \
( \
dma_set_coherent_mask( \
&((pdx)->pPciDevice->dev), \
(mask) \
) \
)
#endif
/***********************************************************
* DMA_BIT_MASK
*
* This macro is used to specify bit masks (e.g dma_set_mask).
* It was introduced in 2.6.24
**********************************************************/
#if LINUX_VERSION_CODE < KERNEL_VERSION(2,6,24)
#define PLX_DMA_BIT_MASK(n) (((n) == 64) ? ~0ULL : ((1ULL<<(n))-1))
#else
#define PLX_DMA_BIT_MASK DMA_BIT_MASK
#endif
#endif // _PLX_SYSDEP_H_