Root/Documentation/bus-virt-phys-mapping.txt

1[ NOTE: The virt_to_bus() and bus_to_virt() functions have been
2    superseded by the functionality provided by the PCI DMA interface
3    (see Documentation/PCI/PCI-DMA-mapping.txt). They continue
4    to be documented below for historical purposes, but new code
5    must not use them. --davidm 00/12/12 ]
6
7[ This is a mail message in response to a query on IO mapping, thus the
8  strange format for a "document" ]
9
10The AHA-1542 is a bus-master device, and your patch makes the driver give the
11controller the physical address of the buffers, which is correct on x86
12(because all bus master devices see the physical memory mappings directly).
13
14However, on many setups, there are actually _three_ different ways of looking
15at memory addresses, and in this case we actually want the third, the
16so-called "bus address".
17
18Essentially, the three ways of addressing memory are (this is "real memory",
19that is, normal RAM--see later about other details):
20
21 - CPU untranslated. This is the "physical" address. Physical address
22   0 is what the CPU sees when it drives zeroes on the memory bus.
23
24 - CPU translated address. This is the "virtual" address, and is
25   completely internal to the CPU itself with the CPU doing the appropriate
26   translations into "CPU untranslated".
27
28 - bus address. This is the address of memory as seen by OTHER devices,
29   not the CPU. Now, in theory there could be many different bus
30   addresses, with each device seeing memory in some device-specific way, but
31   happily most hardware designers aren't actually actively trying to make
32   things any more complex than necessary, so you can assume that all
33   external hardware sees the memory the same way.
34
35Now, on normal PCs the bus address is exactly the same as the physical
36address, and things are very simple indeed. However, they are that simple
37because the memory and the devices share the same address space, and that is
38not generally necessarily true on other PCI/ISA setups.
39
40Now, just as an example, on the PReP (PowerPC Reference Platform), the
41CPU sees a memory map something like this (this is from memory):
42
43    0-2 GB "real memory"
44    2 GB-3 GB "system IO" (inb/out and similar accesses on x86)
45    3 GB-4 GB "IO memory" (shared memory over the IO bus)
46
47Now, that looks simple enough. However, when you look at the same thing from
48the viewpoint of the devices, you have the reverse, and the physical memory
49address 0 actually shows up as address 2 GB for any IO master.
50
51So when the CPU wants any bus master to write to physical memory 0, it
52has to give the master address 0x80000000 as the memory address.
53
54So, for example, depending on how the kernel is actually mapped on the
55PPC, you can end up with a setup like this:
56
57 physical address: 0
58 virtual address: 0xC0000000
59 bus address: 0x80000000
60
61where all the addresses actually point to the same thing. It's just seen
62through different translations..
63
64Similarly, on the Alpha, the normal translation is
65
66 physical address: 0
67 virtual address: 0xfffffc0000000000
68 bus address: 0x40000000
69
70(but there are also Alphas where the physical address and the bus address
71are the same).
72
73Anyway, the way to look up all these translations, you do
74
75    #include <asm/io.h>
76
77    phys_addr = virt_to_phys(virt_addr);
78    virt_addr = phys_to_virt(phys_addr);
79     bus_addr = virt_to_bus(virt_addr);
80    virt_addr = bus_to_virt(bus_addr);
81
82Now, when do you need these?
83
84You want the _virtual_ address when you are actually going to access that
85pointer from the kernel. So you can have something like this:
86
87    /*
88     * this is the hardware "mailbox" we use to communicate with
89     * the controller. The controller sees this directly.
90     */
91    struct mailbox {
92        __u32 status;
93        __u32 bufstart;
94        __u32 buflen;
95        ..
96    } mbox;
97
98        unsigned char * retbuffer;
99
100        /* get the address from the controller */
101        retbuffer = bus_to_virt(mbox.bufstart);
102        switch (retbuffer[0]) {
103            case STATUS_OK:
104                ...
105
106on the other hand, you want the bus address when you have a buffer that
107you want to give to the controller:
108
109    /* ask the controller to read the sense status into "sense_buffer" */
110    mbox.bufstart = virt_to_bus(&sense_buffer);
111    mbox.buflen = sizeof(sense_buffer);
112    mbox.status = 0;
113    notify_controller(&mbox);
114
115And you generally _never_ want to use the physical address, because you can't
116use that from the CPU (the CPU only uses translated virtual addresses), and
117you can't use it from the bus master.
118
119So why do we care about the physical address at all? We do need the physical
120address in some cases, it's just not very often in normal code. The physical
121address is needed if you use memory mappings, for example, because the
122"remap_pfn_range()" mm function wants the physical address of the memory to
123be remapped as measured in units of pages, a.k.a. the pfn (the memory
124management layer doesn't know about devices outside the CPU, so it
125shouldn't need to know about "bus addresses" etc).
126
127NOTE NOTE NOTE! The above is only one part of the whole equation. The above
128only talks about "real memory", that is, CPU memory (RAM).
129
130There is a completely different type of memory too, and that's the "shared
131memory" on the PCI or ISA bus. That's generally not RAM (although in the case
132of a video graphics card it can be normal DRAM that is just used for a frame
133buffer), but can be things like a packet buffer in a network card etc.
134
135This memory is called "PCI memory" or "shared memory" or "IO memory" or
136whatever, and there is only one way to access it: the readb/writeb and
137related functions. You should never take the address of such memory, because
138there is really nothing you can do with such an address: it's not
139conceptually in the same memory space as "real memory" at all, so you cannot
140just dereference a pointer. (Sadly, on x86 it _is_ in the same memory space,
141so on x86 it actually works to just deference a pointer, but it's not
142portable).
143
144For such memory, you can do things like
145
146 - reading:
147    /*
148     * read first 32 bits from ISA memory at 0xC0000, aka
149     * C000:0000 in DOS terms
150     */
151    unsigned int signature = isa_readl(0xC0000);
152
153 - remapping and writing:
154    /*
155     * remap framebuffer PCI memory area at 0xFC000000,
156     * size 1MB, so that we can access it: We can directly
157     * access only the 640k-1MB area, so anything else
158     * has to be remapped.
159     */
160    void __iomem *baseptr = ioremap(0xFC000000, 1024*1024);
161
162    /* write a 'A' to the offset 10 of the area */
163    writeb('A',baseptr+10);
164
165    /* unmap when we unload the driver */
166    iounmap(baseptr);
167
168 - copying and clearing:
169    /* get the 6-byte Ethernet address at ISA address E000:0040 */
170    memcpy_fromio(kernel_buffer, 0xE0040, 6);
171    /* write a packet to the driver */
172    memcpy_toio(0xE1000, skb->data, skb->len);
173    /* clear the frame buffer */
174    memset_io(0xA0000, 0, 0x10000);
175
176OK, that just about covers the basics of accessing IO portably. Questions?
177Comments? You may think that all the above is overly complex, but one day you
178might find yourself with a 500 MHz Alpha in front of you, and then you'll be
179happy that your driver works ;)
180
181Note that kernel versions 2.0.x (and earlier) mistakenly called the
182ioremap() function "vremap()". ioremap() is the proper name, but I
183didn't think straight when I wrote it originally. People who have to
184support both can do something like:
185 
186    /* support old naming silliness */
187    #if LINUX_VERSION_CODE < 0x020100
188    #define ioremap vremap
189    #define iounmap vfree
190    #endif
191 
192at the top of their source files, and then they can use the right names
193even on 2.0.x systems.
194
195And the above sounds worse than it really is. Most real drivers really
196don't do all that complex things (or rather: the complexity is not so
197much in the actual IO accesses as in error handling and timeouts etc).
198It's generally not hard to fix drivers, and in many cases the code
199actually looks better afterwards:
200
201    unsigned long signature = *(unsigned int *) 0xC0000;
202        vs
203    unsigned long signature = readl(0xC0000);
204
205I think the second version actually is more readable, no?
206
207        Linus
208
209

Archive Download this file



interactive