You cannot select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
microwatt/mmu.vhdl

515 lines
18 KiB
VHDL

library ieee;
use ieee.std_logic_1164.all;
use ieee.numeric_std.all;
library work;
use work.common.all;
-- Radix MMU
-- Supports 4-level trees as in arch 3.0B, but not the two-step translation for
-- guests under a hypervisor (i.e. there is no gRA -> hRA translation).
entity mmu is
port (
clk : in std_ulogic;
rst : in std_ulogic;
l_in : in Loadstore1ToMmuType;
l_out : out MmuToLoadstore1Type;
d_out : out MmuToDcacheType;
Add TLB to icache This adds a direct-mapped TLB to the icache, with 64 entries by default. Execute1 now sends a "virt_mode" signal from MSR[IR] to fetch1 along with redirects to indicate whether instruction addresses should be translated through the TLB, and fetch1 sends that on to icache. Similarly a "priv_mode" signal is sent to indicate the privilege mode for instruction fetches. This means that changes to MSR[IR] or MSR[PR] don't take effect until the next redirect, meaning an isync, rfid, branch, etc. The icache uses a hash of the effective address (i.e. next instruction address) to index the TLB. The hash is an XOR of three fields of the address; with a 64-entry TLB, the fields are bits 12--17, 18--23 and 24--29 of the address. TLB invalidations simply invalidate the indexed TLB entry without checking the contents. If the icache detects a TLB miss with virt_mode=1, it will send a fetch_failed indication through fetch2 to decode1, which will turn it into a special OP_FETCH_FAILED opcode with unit=LDST. That will get sent down to loadstore1 which will currently just raise a Instruction Storage Interrupt (0x400) exception. One bit in the PTE obtained from the TLB is used to check whether an instruction access is allowed -- the privilege bit (bit 3). If bit 3 is 1 and priv_mode=0, then a fetch_failed indication is sent down to fetch2 and to decode1, which generates an OP_FETCH_FAILED. Any PTEs with PTE bit 0 (EAA[3]) clear or bit 8 (R) clear should not be put into the iTLB since such PTEs would not allow execution by any context. Tlbie operations get sent from mmu to icache over a new connection. Unfortunately the privileged instruction tests are broken for now. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
d_in : in DcacheToMmuType;
i_out : out MmuToIcacheType
);
end mmu;
architecture behave of mmu is
type state_t is (IDLE,
DO_TLBIE,
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
TLB_WAIT,
PART_TBL_READ,
PART_TBL_WAIT,
PART_TBL_DONE,
PROC_TBL_READ,
PROC_TBL_WAIT,
SEGMENT_CHECK,
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
RADIX_LOOKUP,
RADIX_READ_WAIT,
RADIX_LOAD_TLB,
RADIX_FINISH
);
type reg_stage_t is record
-- latched request from loadstore1
valid : std_ulogic;
iside : std_ulogic;
store : std_ulogic;
priv : std_ulogic;
addr : std_ulogic_vector(63 downto 0);
inval_all : std_ulogic;
-- config SPRs
ptcr : std_ulogic_vector(63 downto 0);
pid : std_ulogic_vector(31 downto 0);
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
-- internal state
state : state_t;
done : std_ulogic;
err : std_ulogic;
prtbl : std_ulogic_vector(63 downto 0);
ptb_valid : std_ulogic;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
pgtbl0 : std_ulogic_vector(63 downto 0);
pt0_valid : std_ulogic;
pgtbl3 : std_ulogic_vector(63 downto 0);
pt3_valid : std_ulogic;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
shift : unsigned(5 downto 0);
mask_size : unsigned(4 downto 0);
pgbase : std_ulogic_vector(55 downto 0);
pde : std_ulogic_vector(63 downto 0);
invalid : std_ulogic;
badtree : std_ulogic;
segerror : std_ulogic;
perm_err : std_ulogic;
rc_error : std_ulogic;
end record;
signal r, rin : reg_stage_t;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
signal addrsh : std_ulogic_vector(15 downto 0);
signal mask : std_ulogic_vector(15 downto 0);
signal finalmask : std_ulogic_vector(43 downto 0);
begin
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
-- Multiplex internal SPR values back to loadstore1, selected
-- by l_in.sprn.
l_out.sprval <= r.ptcr when l_in.sprn(8) = '1' else x"00000000" & r.pid;
mmu_0: process(clk)
begin
if rising_edge(clk) then
if rst = '1' then
r.state <= IDLE;
r.valid <= '0';
r.ptb_valid <= '0';
r.pt0_valid <= '0';
r.pt3_valid <= '0';
r.ptcr <= (others => '0');
r.pid <= (others => '0');
else
if rin.valid = '1' then
report "MMU got tlb miss for " & to_hstring(rin.addr);
end if;
if l_out.done = '1' then
report "MMU completing op without error";
end if;
if l_out.err = '1' then
report "MMU completing op with err invalid=" & std_ulogic'image(l_out.invalid) &
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
" badtree=" & std_ulogic'image(l_out.badtree);
end if;
if rin.state = RADIX_LOOKUP then
report "radix lookup shift=" & integer'image(to_integer(rin.shift)) &
" msize=" & integer'image(to_integer(rin.mask_size));
end if;
if r.state = RADIX_LOOKUP then
report "send load addr=" & to_hstring(d_out.addr) &
" addrsh=" & to_hstring(addrsh) & " mask=" & to_hstring(mask);
end if;
r <= rin;
end if;
end if;
end process;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
-- Shift address bits 61--12 right by 0--47 bits and
-- supply the least significant 16 bits of the result.
addrshifter: process(all)
variable sh1 : std_ulogic_vector(30 downto 0);
variable sh2 : std_ulogic_vector(18 downto 0);
variable result : std_ulogic_vector(15 downto 0);
begin
case r.shift(5 downto 4) is
when "00" =>
sh1 := r.addr(42 downto 12);
when "01" =>
sh1 := r.addr(58 downto 28);
when others =>
sh1 := "0000000000000" & r.addr(61 downto 44);
end case;
case r.shift(3 downto 2) is
when "00" =>
sh2 := sh1(18 downto 0);
when "01" =>
sh2 := sh1(22 downto 4);
when "10" =>
sh2 := sh1(26 downto 8);
when others =>
sh2 := sh1(30 downto 12);
end case;
case r.shift(1 downto 0) is
when "00" =>
result := sh2(15 downto 0);
when "01" =>
result := sh2(16 downto 1);
when "10" =>
result := sh2(17 downto 2);
when others =>
result := sh2(18 downto 3);
end case;
addrsh <= result;
end process;
-- generate mask for extracting address fields for PTE address generation
addrmaskgen: process(all)
variable m : std_ulogic_vector(15 downto 0);
begin
-- mask_count has to be >= 5
m := x"001f";
for i in 5 to 15 loop
if i < to_integer(r.mask_size) then
m(i) := '1';
end if;
end loop;
mask <= m;
end process;
-- generate mask for extracting address bits to go in TLB entry
-- in order to support pages > 4kB
finalmaskgen: process(all)
variable m : std_ulogic_vector(43 downto 0);
begin
m := (others => '0');
for i in 0 to 43 loop
if i < to_integer(r.shift) then
m(i) := '1';
end if;
end loop;
finalmask <= m;
end process;
mmu_1: process(all)
variable v : reg_stage_t;
variable dcreq : std_ulogic;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
variable tlb_load : std_ulogic;
variable itlb_load : std_ulogic;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
variable tlbie_req : std_ulogic;
variable ptbl_rd : std_ulogic;
variable prtbl_rd : std_ulogic;
variable pt_valid : std_ulogic;
variable effpid : std_ulogic_vector(31 downto 0);
variable prtable_addr : std_ulogic_vector(63 downto 0);
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
variable rts : unsigned(5 downto 0);
variable mbits : unsigned(5 downto 0);
variable pgtable_addr : std_ulogic_vector(63 downto 0);
variable pte : std_ulogic_vector(63 downto 0);
variable tlb_data : std_ulogic_vector(63 downto 0);
variable nonzero : std_ulogic;
variable pgtbl : std_ulogic_vector(63 downto 0);
variable perm_ok : std_ulogic;
variable rc_ok : std_ulogic;
variable addr : std_ulogic_vector(63 downto 0);
variable data : std_ulogic_vector(63 downto 0);
begin
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
v := r;
v.valid := '0';
dcreq := '0';
v.done := '0';
v.err := '0';
v.invalid := '0';
v.badtree := '0';
v.segerror := '0';
v.perm_err := '0';
v.rc_error := '0';
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
tlb_load := '0';
itlb_load := '0';
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
tlbie_req := '0';
v.inval_all := '0';
ptbl_rd := '0';
prtbl_rd := '0';
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
-- Radix tree data structures in memory are big-endian,
-- so we need to byte-swap them
for i in 0 to 7 loop
data(i * 8 + 7 downto i * 8) := d_in.data((7 - i) * 8 + 7 downto (7 - i) * 8);
end loop;
case r.state is
when IDLE =>
if l_in.addr(63) = '0' then
pgtbl := r.pgtbl0;
pt_valid := r.pt0_valid;
else
pgtbl := r.pgtbl3;
pt_valid := r.pt3_valid;
end if;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
-- rts == radix tree size, # address bits being translated
rts := unsigned('0' & pgtbl(62 downto 61) & pgtbl(7 downto 5));
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
-- mbits == # address bits to index top level of tree
mbits := unsigned('0' & pgtbl(4 downto 0));
-- set v.shift to rts so that we can use finalmask for the segment check
v.shift := rts;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
v.mask_size := mbits(4 downto 0);
v.pgbase := pgtbl(55 downto 8) & x"00";
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
if l_in.valid = '1' then
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
v.addr := l_in.addr;
v.iside := l_in.iside;
v.store := not (l_in.load or l_in.iside);
v.priv := l_in.priv;
if l_in.tlbie = '1' then
-- Invalidate all iTLB/dTLB entries for tlbie with
-- RB[IS] != 0 or RB[AP] != 0, or for slbia
v.inval_all := l_in.slbia or l_in.addr(11) or l_in.addr(10) or
l_in.addr(7) or l_in.addr(6) or l_in.addr(5);
-- The RIC field of the tlbie instruction comes across on the
-- sprn bus as bits 2--3. RIC=2 flushes process table caches.
if l_in.sprn(3) = '1' then
v.pt0_valid := '0';
v.pt3_valid := '0';
v.ptb_valid := '0';
end if;
v.state := DO_TLBIE;
else
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
v.valid := '1';
if r.ptb_valid = '0' then
-- need to fetch process table base from partition table
v.state := PART_TBL_READ;
elsif pt_valid = '0' then
-- need to fetch process table entry
-- set v.shift so we can use finalmask for generating
-- the process table entry address
v.shift := unsigned('0' & r.prtbl(4 downto 0));
v.state := PROC_TBL_READ;
elsif mbits = 0 then
-- Use RPDS = 0 to disable radix tree walks
v.state := RADIX_FINISH;
v.invalid := '1';
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
else
v.state := SEGMENT_CHECK;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
end if;
end if;
end if;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
if l_in.mtspr = '1' then
-- Move to PID needs to invalidate L1 TLBs and cached
-- pgtbl0 value. Move to PTCR does that plus
-- invalidating the cached pgtbl3 and prtbl values as well.
if l_in.sprn(8) = '0' then
v.pid := l_in.rs(31 downto 0);
else
v.ptcr := l_in.rs;
v.pt3_valid := '0';
v.ptb_valid := '0';
end if;
v.pt0_valid := '0';
v.inval_all := '1';
v.state := DO_TLBIE;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
end if;
when DO_TLBIE =>
dcreq := '1';
tlbie_req := '1';
v.state := TLB_WAIT;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
when TLB_WAIT =>
if d_in.done = '1' then
v.state := RADIX_FINISH;
end if;
when PART_TBL_READ =>
dcreq := '1';
ptbl_rd := '1';
v.state := PART_TBL_WAIT;
when PART_TBL_WAIT =>
if d_in.done = '1' then
v.prtbl := data;
v.ptb_valid := '1';
v.state := PART_TBL_DONE;
end if;
when PART_TBL_DONE =>
v.shift := unsigned('0' & r.prtbl(4 downto 0));
v.state := PROC_TBL_READ;
when PROC_TBL_READ =>
dcreq := '1';
prtbl_rd := '1';
v.state := PROC_TBL_WAIT;
when PROC_TBL_WAIT =>
if d_in.done = '1' then
if r.addr(63) = '1' then
v.pgtbl3 := data;
v.pt3_valid := '1';
else
v.pgtbl0 := data;
v.pt0_valid := '1';
end if;
-- rts == radix tree size, # address bits being translated
rts := unsigned('0' & data(62 downto 61) & data(7 downto 5));
-- mbits == # address bits to index top level of tree
mbits := unsigned('0' & data(4 downto 0));
-- set v.shift to rts so that we can use finalmask for the segment check
v.shift := rts;
v.mask_size := mbits(4 downto 0);
v.pgbase := data(55 downto 8) & x"00";
if mbits = 0 then
v.state := RADIX_FINISH;
v.invalid := '1';
else
v.state := SEGMENT_CHECK;
end if;
end if;
if d_in.err = '1' then
v.state := RADIX_FINISH;
v.badtree := '1';
end if;
when SEGMENT_CHECK =>
mbits := '0' & r.mask_size;
v.shift := r.shift + (31 - 12) - mbits;
nonzero := or(r.addr(61 downto 31) and not finalmask(30 downto 0));
if r.addr(63) /= r.addr(62) or nonzero = '1' then
v.state := RADIX_FINISH;
v.segerror := '1';
elsif mbits < 5 or mbits > 16 or mbits > (r.shift + (31 - 12)) then
v.state := RADIX_FINISH;
v.badtree := '1';
else
v.state := RADIX_LOOKUP;
end if;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
when RADIX_LOOKUP =>
dcreq := '1';
v.state := RADIX_READ_WAIT;
when RADIX_READ_WAIT =>
if d_in.done = '1' then
v.pde := data;
-- test valid bit
if data(63) = '1' then
-- test leaf bit
if data(62) = '1' then
-- check permissions and RC bits
perm_ok := '0';
if r.priv = '1' or data(3) = '0' then
if r.iside = '0' then
perm_ok := data(1) or (data(2) and not r.store);
else
-- no IAMR, so no KUEP support for now
-- deny execute permission if cache inhibited
perm_ok := data(0) and not data(5);
end if;
end if;
rc_ok := data(8) and (data(7) or not r.store);
if perm_ok = '1' and rc_ok = '1' then
v.state := RADIX_LOAD_TLB;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
else
v.state := RADIX_FINISH;
v.perm_err := not perm_ok;
-- permission error takes precedence over RC error
v.rc_error := perm_ok;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
end if;
else
mbits := unsigned('0' & data(4 downto 0));
if mbits < 5 or mbits > 16 or mbits > r.shift then
v.state := RADIX_FINISH;
v.badtree := '1';
else
v.shift := v.shift - mbits;
v.mask_size := mbits(4 downto 0);
v.pgbase := data(55 downto 8) & x"00";
v.state := RADIX_LOOKUP;
end if;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
end if;
else
-- non-present PTE, generate a DSI
v.state := RADIX_FINISH;
v.invalid := '1';
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
end if;
end if;
if d_in.err = '1' then
v.state := RADIX_FINISH;
v.badtree := '1';
end if;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
when RADIX_LOAD_TLB =>
tlb_load := '1';
if r.iside = '0' then
dcreq := '1';
v.state := TLB_WAIT;
else
itlb_load := '1';
v.state := IDLE;
end if;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
when RADIX_FINISH =>
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
v.state := IDLE;
end case;
if v.state = RADIX_FINISH or (v.state = RADIX_LOAD_TLB and r.iside = '1') then
v.err := v.invalid or v.badtree or v.segerror or v.perm_err or v.rc_error;
v.done := not v.err;
end if;
if r.addr(63) = '1' then
effpid := x"00000000";
else
effpid := r.pid;
end if;
prtable_addr := x"00" & r.prtbl(55 downto 36) &
((r.prtbl(35 downto 12) and not finalmask(23 downto 0)) or
(effpid(31 downto 8) and finalmask(23 downto 0))) &
effpid(7 downto 0) & "0000";
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
pgtable_addr := x"00" & r.pgbase(55 downto 19) &
((r.pgbase(18 downto 3) and not mask) or (addrsh and mask)) &
"000";
pte := x"00" &
((r.pde(55 downto 12) and not finalmask) or (r.addr(55 downto 12) and finalmask))
& r.pde(11 downto 0);
-- update registers
rin <= v;
-- drive outputs
if tlbie_req = '1' then
addr := r.addr;
tlb_data := (others => '0');
elsif tlb_load = '1' then
addr := r.addr(63 downto 12) & x"000";
tlb_data := pte;
elsif ptbl_rd = '1' then
addr := x"00" & r.ptcr(55 downto 12) & x"008";
tlb_data := (others => '0');
elsif prtbl_rd = '1' then
addr := prtable_addr;
tlb_data := (others => '0');
else
addr := pgtable_addr;
tlb_data := (others => '0');
end if;
l_out.done <= r.done;
l_out.err <= r.err;
l_out.invalid <= r.invalid;
l_out.badtree <= r.badtree;
l_out.segerr <= r.segerror;
l_out.perm_error <= r.perm_err;
l_out.rc_error <= r.rc_error;
d_out.valid <= dcreq;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
d_out.tlbie <= tlbie_req;
d_out.doall <= r.inval_all;
MMU: Implement radix page table machinery This adds the necessary machinery to the MMU for it to do radix page table walks. The core elements are a shifter that can shift the address right by between 0 and 47 bits, a mask generator that can generate a mask of between 5 and 16 bits, a final mask generator, and new states in the state machine. (The final mask generator is used for transferring bits of the original address into the resulting TLB entry when the leaf PTE corresponds to a page size larger than 4kB.) The hardware does not implement a partition table or a process table. Software is expected to load the appropriate process table entry into a new SPR called PGTBL0, SPR 720. The contents should be formatted as described in Book III section 5.7.6.2 of the Power ISA v3.0B. PGTBL0 is set to 0 on hard reset. At present, the top two bits of the address (the quadrant) are ignored. There is currently no caching of any step in the translation process or of the final result, other than the entry created in the dTLB. That entry is a 4k page entry even if the leaf PTE found in the walk corresponds to a larger page size. This implementation can handle almost any page table layout and any page size. The RTS field (in PGTBL0) can have any value between 0 and 31, corresponding to a total address space size between 2^31 and 2^62 bytes. The RPDS field of PGTBL0 can be any value between 5 and 16, except that a value of 0 is taken to disable radix page table walking (for use when one is using software loading of TLB entries). The NLS field of the page directory entries can have any value between 5 and 16. The minimum page size is 4kB, meaning that the sum of RPDS and the NLS values of the PDEs found on the path to a leaf PTE must be less than or equal to RTS + 31 - 12. The PGTBL0 SPR is in the mmu module; thus this adds a path for loadstore1 to read and write SPRs in mmu. This adds code in dcache to service doubleword read requests from the MMU, as well as requests to write dTLB entries. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
d_out.tlbld <= tlb_load;
d_out.addr <= addr;
d_out.pte <= tlb_data;
Add TLB to icache This adds a direct-mapped TLB to the icache, with 64 entries by default. Execute1 now sends a "virt_mode" signal from MSR[IR] to fetch1 along with redirects to indicate whether instruction addresses should be translated through the TLB, and fetch1 sends that on to icache. Similarly a "priv_mode" signal is sent to indicate the privilege mode for instruction fetches. This means that changes to MSR[IR] or MSR[PR] don't take effect until the next redirect, meaning an isync, rfid, branch, etc. The icache uses a hash of the effective address (i.e. next instruction address) to index the TLB. The hash is an XOR of three fields of the address; with a 64-entry TLB, the fields are bits 12--17, 18--23 and 24--29 of the address. TLB invalidations simply invalidate the indexed TLB entry without checking the contents. If the icache detects a TLB miss with virt_mode=1, it will send a fetch_failed indication through fetch2 to decode1, which will turn it into a special OP_FETCH_FAILED opcode with unit=LDST. That will get sent down to loadstore1 which will currently just raise a Instruction Storage Interrupt (0x400) exception. One bit in the PTE obtained from the TLB is used to check whether an instruction access is allowed -- the privilege bit (bit 3). If bit 3 is 1 and priv_mode=0, then a fetch_failed indication is sent down to fetch2 and to decode1, which generates an OP_FETCH_FAILED. Any PTEs with PTE bit 0 (EAA[3]) clear or bit 8 (R) clear should not be put into the iTLB since such PTEs would not allow execution by any context. Tlbie operations get sent from mmu to icache over a new connection. Unfortunately the privileged instruction tests are broken for now. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
i_out.tlbld <= itlb_load;
Add TLB to icache This adds a direct-mapped TLB to the icache, with 64 entries by default. Execute1 now sends a "virt_mode" signal from MSR[IR] to fetch1 along with redirects to indicate whether instruction addresses should be translated through the TLB, and fetch1 sends that on to icache. Similarly a "priv_mode" signal is sent to indicate the privilege mode for instruction fetches. This means that changes to MSR[IR] or MSR[PR] don't take effect until the next redirect, meaning an isync, rfid, branch, etc. The icache uses a hash of the effective address (i.e. next instruction address) to index the TLB. The hash is an XOR of three fields of the address; with a 64-entry TLB, the fields are bits 12--17, 18--23 and 24--29 of the address. TLB invalidations simply invalidate the indexed TLB entry without checking the contents. If the icache detects a TLB miss with virt_mode=1, it will send a fetch_failed indication through fetch2 to decode1, which will turn it into a special OP_FETCH_FAILED opcode with unit=LDST. That will get sent down to loadstore1 which will currently just raise a Instruction Storage Interrupt (0x400) exception. One bit in the PTE obtained from the TLB is used to check whether an instruction access is allowed -- the privilege bit (bit 3). If bit 3 is 1 and priv_mode=0, then a fetch_failed indication is sent down to fetch2 and to decode1, which generates an OP_FETCH_FAILED. Any PTEs with PTE bit 0 (EAA[3]) clear or bit 8 (R) clear should not be put into the iTLB since such PTEs would not allow execution by any context. Tlbie operations get sent from mmu to icache over a new connection. Unfortunately the privileged instruction tests are broken for now. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
i_out.tlbie <= tlbie_req;
i_out.doall <= r.inval_all;
i_out.addr <= addr;
i_out.pte <= tlb_data;
Add TLB to icache This adds a direct-mapped TLB to the icache, with 64 entries by default. Execute1 now sends a "virt_mode" signal from MSR[IR] to fetch1 along with redirects to indicate whether instruction addresses should be translated through the TLB, and fetch1 sends that on to icache. Similarly a "priv_mode" signal is sent to indicate the privilege mode for instruction fetches. This means that changes to MSR[IR] or MSR[PR] don't take effect until the next redirect, meaning an isync, rfid, branch, etc. The icache uses a hash of the effective address (i.e. next instruction address) to index the TLB. The hash is an XOR of three fields of the address; with a 64-entry TLB, the fields are bits 12--17, 18--23 and 24--29 of the address. TLB invalidations simply invalidate the indexed TLB entry without checking the contents. If the icache detects a TLB miss with virt_mode=1, it will send a fetch_failed indication through fetch2 to decode1, which will turn it into a special OP_FETCH_FAILED opcode with unit=LDST. That will get sent down to loadstore1 which will currently just raise a Instruction Storage Interrupt (0x400) exception. One bit in the PTE obtained from the TLB is used to check whether an instruction access is allowed -- the privilege bit (bit 3). If bit 3 is 1 and priv_mode=0, then a fetch_failed indication is sent down to fetch2 and to decode1, which generates an OP_FETCH_FAILED. Any PTEs with PTE bit 0 (EAA[3]) clear or bit 8 (R) clear should not be put into the iTLB since such PTEs would not allow execution by any context. Tlbie operations get sent from mmu to icache over a new connection. Unfortunately the privileged instruction tests are broken for now. Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
5 years ago
end process;
end;