Paul Mackerras [Mon, 13 Jul 2020 07:43:52 +0000 (17:43 +1000)]
loadstore1: Separate address calculation for MMU to ease timing
This computes the address sent to the MMU separately from that sent
to the dcache. This means that the address sent to the MMU doesn't
have the delay through the lsu_sum adder, making it available earlier.
The path through the lsu_sum adder and through the MMU to the MMU
done and err outputs showed up as a critical path on some builds.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Mon, 13 Jul 2020 02:18:53 +0000 (12:18 +1000)]
loadstore1: Generate busy signal earlier
This makes the calculation of busy as simple as possible and dependent
only on register outputs. The timing of busy is critical, as it gates
the valid signal for the next instruction, and therefore any delays
in dropping busy at the end of a load or store directly impact the
timing of a host of other paths.
This also separates the 'done without error' and 'done with error'
cases from the MMU into separate signals that are both driven directly
from registers.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sun, 12 Jul 2020 10:05:53 +0000 (20:05 +1000)]
dcache: Output separate done-without-error and error-done signals
This reduces the complexity of the logic in the places where these
signals are used.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 11 Jul 2020 12:23:31 +0000 (22:23 +1000)]
dcache: Ease timing on calculation of acks remaining
This moves the incrementing or decrementing of r1.acks_pending
to the cycle after a strobe is output or an ack is seen on the
wishbone, and simplifies the logic that determines whether the
cycle is now complete. This means that the path from seeing
req_op equal to OP_STORE_HIT or OP_STORE_MISS to setting r1.state
and r1.cyc now just involves the stbs_done bit rather than a more
complex calculation involving the possibly incremented r1.acks_pending.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 11 Jul 2020 07:46:03 +0000 (17:46 +1000)]
dcache: Improve timing of valid/done outputs
This makes d_out.valid and m_out.done come directly from registers in
order to improve timing. The inputs to the registers are set by the
same conditions that cause r1.hit_load_valid, r1.slow_valid,
r1.error_done and r1.stcx_fail to be set.
Note that the STORE_WAIT_ACK state doesn't test r1.mmu_req but assumes
that the request came from loadstore1. This is because we normally
have r1.full = 0 in this state, which means that r1.mmu_req can
change at any time.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 11 Jul 2020 05:40:27 +0000 (15:40 +1000)]
core: Don't generate logic for log data when LOG_LENGTH = 0
This adds "if LOG_LENGTH > 0 generate" to the places in the core
where log output data is latched, so that when LOG_LENGTH = 0 we
don't create the logic to collect the data which won't be stored.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 11 Jul 2020 02:05:43 +0000 (12:05 +1000)]
countzero: Faster algorithm for count leading/trailing zeroes
This uses an algorithm for count leading/trailing zeroes that is
faster on FPGAs, which makes timing easier. cntlz* and cnttz*
still take two cycles, though.
For count trailing zeroes, we compute x & -x, which for non-zero x
has a single 1 bit in the position of the least-significant 1 bit
in x. This one-hot representation can then be converted to a bit
number with six 32-input OR gates. For count leading zeroes, we
simply do a bit-reversal on x and then use the same algorithm.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 11 Jul 2020 01:00:53 +0000 (11:00 +1000)]
MMU: Improve timing of done signal back to loadstore1
This makes the l_out.done signal come from a clean latch, which
improves timing. The cost is that TLB load and invalidation
operations to the dcache now signal done back to loadstore1 one
cycle later than before, but that doesn't seem to affect overall
performance noticeably.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 10 Jul 2020 23:10:24 +0000 (09:10 +1000)]
dcache: Remove dependency of r1.wb.adr/dat/sel on req_op
This improves timing by setting r1.wb.{adr,dat,sel} to the next
request when doing a write cycle on the wishbone before we know
whether the next request has a TLB and cache hit or not, i.e.
without depending on req_op. r1.wb.stb still depends on req_op.
This contains a workaround for what is probably a bug elsewhere,
in that changing r1.wb.sel unconditionally once we see stall=0
from the wishbone causes incorrect behaviour. Making it
conditional on there being a valid following request appears
to fix the problem.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 10 Jul 2020 10:32:35 +0000 (20:32 +1000)]
dcache: Update TLB PLRU one cycle later
This puts the inputs to the TLB PLRU through a register stage, so
the TLB PLRU update is done in the cycle after the TLB tag
matching rather than the same cycle. This improves timing.
The PLRU output is only used when writing the TLB in response to
a tlbwe request from the MMU, and that doesn't happen within one
cycle of a virtual-mode load or store, so the fact that the
tlb victim way information is delayed by one cycle doesn't
create any problems.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 10 Jul 2020 10:11:15 +0000 (20:11 +1000)]
loadstore1: Eliminate two_dwords variable
The computation of two_dwords from r.second_bytes has shown up as
part of a critical path at times. Instead we add a 'last_dword'
flag to the reg_stage_t record which tells us more directly
whether a valid flag coming in from dcache means that the
instruction is done, thereby shortening the path to the busy output
back to execute1.
This also simplifies some of the trim_ctl logic. The two_dwords = 0
case could never have use_second(i) = 1 for any of the bytes being
transferred, so "not use_second(i)" is always 1.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 10 Jul 2020 09:07:47 +0000 (19:07 +1000)]
execute1: Ease timing on redirect_nia
This eliminates a dependency of r.f.redirect_nia on the carry out
from the main adder in the case of a conditional trap instruction.
We can set r.f.redirect_nia unconditionally, even if no interrupt
is generated.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 10 Jul 2020 07:47:52 +0000 (17:47 +1000)]
dcache: Do PLRU update one cycle later
This does the PLRU update based on r1.cache_hit and r1.hit_way rather
than req_op and req_hit_way, which means there is now a register
between the TLB and cache tag lookup and the PLRU update, which should
help with timing.
The PLRU victim selection now becomes valid one cycle later, in the
cycle where r1.write_tag = 1. We now have replace_way coming from
the PLRU when r1.write_tag = 1 and from r1.store_way at other times,
and we use that instead of r1.store_way in situations where we need
it to be valid in the first cycle of the RELOAD_WAIT_ACK state.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 10 Jul 2020 00:04:56 +0000 (10:04 +1000)]
icache: Do PLRU update one cycle later
This does the PLRU update based on r.hit_valid and r.hit_way rather
than req_is_hit and req_hit_way, which means there is now a register
between the TLB and cache tag lookup and the PLRU update, which
should help with timing.
As a result, the PLRU victim way selection becomes valid one cycle
later, in the cycle when r.state = CLR_TAG. So we have to use the
PLRU output directly in the CLR_TAG state and r.store_way in the
WAIT_ACK state.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Michael Neuling [Thu, 9 Jul 2020 02:25:50 +0000 (12:25 +1000)]
Merge pull request #228 from ozbenh/misc
Misc nexys video fixes
Michael Neuling [Thu, 9 Jul 2020 01:10:45 +0000 (11:10 +1000)]
Merge pull request #222 from iamjpn/master
core: Implement PVR register
Jordan Niethe [Wed, 8 Jul 2020 04:34:42 +0000 (14:34 +1000)]
tests: Add tests for the PVR
The PVR is a privileged read-only SPR. Test reading and writing in both
supervisor and problem state. In supervisor state reading returns
microwatt's assigned PVR number and writing is a noop. In problem state
both reading and writing cause privileged instruction interrupts.
Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
Benjamin Herrenschmidt [Wed, 8 Jul 2020 04:00:27 +0000 (14:00 +1000)]
corefile/nexys_video: Parameter fixes
This fixes up a few issues with parameters:
Only arty has "has_uart1" since we haven't added plumbing for a second UART
anywhere else. Also "uart_is_16550" was mixing on one of the nexys_video
targets, and nexys_video toplevel was missing LOG_LENGTH.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Fri, 26 Jun 2020 13:34:14 +0000 (23:34 +1000)]
fpga: nexys-video: Wire up core_alt_reset
It looks like we left it dangling
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 25 Jun 2020 04:05:03 +0000 (14:05 +1000)]
nexys_video: Fix nexys-video build
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Wed, 8 Jul 2020 06:13:27 +0000 (16:13 +1000)]
spi: Send dummy clocks at boot
When using an FPGA which routes the SPI clock via STARTUPE2 as is
done on the Nexys Video (or optionally on Arty), the HW needs at
least 3 beats of that clock to complete the switch from the internal
config clock to the one we provide.
This works around it by having the SPI controller send 8 dummy
clocks at boot time with CS held high.
Without this, flash identification will fail those boards
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Paul Mackerras [Wed, 8 Jul 2020 00:48:26 +0000 (10:48 +1000)]
Merge pull request #223 from mikey/ecp5
Make ECP5 devices work and add github artifacts
Michael Neuling [Tue, 7 Jul 2020 11:18:34 +0000 (21:18 +1000)]
Create github artifacts for ECP5 devices
ECP5 eval board (tested and working) and Orange Crap (untested)
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Tue, 7 Jul 2020 11:15:34 +0000 (21:15 +1000)]
Add PLL for ECP5 device
Means we can synthesize at 40Mhz (where we currently make timing) and
our UART still works at 115200 baud.
Tested working hello world unmodified with ECP5 eval board. Orange
Crab is updated but is untested.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Anton Blanchard [Tue, 7 Jul 2020 10:54:06 +0000 (20:54 +1000)]
Merge pull request #220 from mikey/ghdl-makefile
Use $(GHDL) rather than ghdl in Makefile
Anton Blanchard [Tue, 7 Jul 2020 10:44:02 +0000 (20:44 +1000)]
Merge pull request #209 from mikey/yosys
Make yosys/nextpnr work and add to CI
Jordan Niethe [Tue, 7 Jul 2020 10:37:52 +0000 (20:37 +1000)]
core: Implement PVR register
Microwatt has been allocated a PVR version of 0x0063. Implement a PVR
with this value.
Signed-off-by: Jordan Niethe <jniethe5@gmail.com>
Michael Neuling [Sat, 4 Jul 2020 01:12:05 +0000 (11:12 +1000)]
Use $(GHDL) rather than ghdl in Makefile
Suggestion from @eine in PR #219.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Mon, 22 Jun 2020 03:10:13 +0000 (13:10 +1000)]
Add yosys/nextpnr ecp5 and verilog build to CI
This works now, so let's make sure it continues to.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Tue, 23 Jun 2020 07:05:23 +0000 (17:05 +1000)]
Add FPGA_TARGET=ECP5-EVN make option for synthesis build
This allows these targets
FPGA_TARGET=ORANGE-CRAB make microwatt.bit
FPGA_TARGET=ECP5-EVN make microwatt.bit
Default is ORANGE-CRAB as before
ECP5-EVN is tested on real hardware. The console only works at 38400 so
needs this in console.c and a recompile of hello_world to work:
-#define UART_FREQ 115200
+#define UART_FREQ 38400
With this 'FPGA_TARGET=ECP5-EVN make prog' works on the ECP5 dev board.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Thu, 2 Jul 2020 04:36:14 +0000 (14:36 +1000)]
Add SYNTH_ECP5_FLAGS option for building
This is useful to specify "-noflatten" which helps CI stay under 8GB
limit.
Normally the AUTONAME stage of yosys will take around 10GB if
operating on the whole design. With -noflatten, AUTONAME occurs only
per VHDL entity, so only consumes around 3GB of memory. This gets us
under the limitations on github actions.
More discussion here:
https://github.com/antonblanchard/microwatt/pull/209#issuecomment-
652186078
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Thu, 2 Jul 2020 05:55:30 +0000 (15:55 +1000)]
Add ram file to synthesis build dependencies
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Thu, 2 Jul 2020 04:34:43 +0000 (14:34 +1000)]
Add uart16550 files to yosys/nextpnr build
These are verilog so need passed to yosys differently than the VHDL
files.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Thu, 2 Jul 2020 04:11:16 +0000 (14:11 +1000)]
Add uart16550 files from fusesoc
These are needed for synthesis that doesn't use fusesoc natively.
These were pulled in via 'fusesoc fetch ::uart16550:1.5.5-r1'
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Mon, 22 Jun 2020 03:09:09 +0000 (13:09 +1000)]
Build to tmp file so nextpnr errors don't confuse make
nextpnr will leave an output file around even when it errors out, so
build to a tmp file and move it when we succeed so we don't confuse
make.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Mon, 22 Jun 2020 03:09:09 +0000 (13:09 +1000)]
Fix building with yosys/nextpnr
Add --no-formal so that asserts are removed by yosys as nextpnr
doesn't like them.
This was suggested by @tgingold here:
https://github.com/YosysHQ/yosys/issues/2068#issuecomment-
644545863
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Mon, 22 Jun 2020 03:11:03 +0000 (13:11 +1000)]
Add yosys builds files to gitignore
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Tue, 23 Jun 2020 07:07:12 +0000 (17:07 +1000)]
Send line feed if we get a carriage return in hello world.
Signed-off-by: Michael Neuling <mikey@neuling.org>
Michael Neuling [Tue, 30 Jun 2020 05:47:36 +0000 (15:47 +1000)]
Merge pull request #216 from paulusmack/cfar
Timing and speed improvements, implement CFAR register
Paul Mackerras [Tue, 30 Jun 2020 05:01:06 +0000 (15:01 +1000)]
Merge pull request #206 from Jbalkind/icachecleanup
Icache constants cleanup
Paul Mackerras [Fri, 19 Jun 2020 10:00:16 +0000 (20:00 +1000)]
execute1: Do forwarding of the CR result to the next instruction
This adds a path to allow the CR result of one instruction to be
forwarded to the next instruction, so that sequences such as
cmp; bc can avoid having a 1-cycle bubble.
Forwarding is not available for dot-form (Rc=1) instructions,
since the CR result for them is calculated in writeback. The
decode.output_cr field is used to identify those instructions
that compute the CR result in execute1.
For some reason, the multiply instructions incorrectly had
output_cr = 1 in the decode tables. This fixes that.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 19 Jun 2020 08:00:37 +0000 (18:00 +1000)]
execute1: Add latch to redirect path
This latches the redirect signal inside execute1, so that it is sent
a cycle later to fetch1 (and to decode/icache as flush). This breaks
a long combinatorial chain from the branch and interrupt detection
in execute1 through the redirect/flush signals all the way back to
fetch1, icache and decode.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 19 Jun 2020 07:13:06 +0000 (17:13 +1000)]
logical: Only do output inversion for OP_AND, OP_OR and OP_XOR
It's not needed for the other ops (popcnt, parity, etc.) and the
logical unit shows up as a critical path from time to time.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Mon, 15 Jun 2020 07:45:55 +0000 (17:45 +1000)]
core: Implement CFAR register
This implements the CFAR SPR as a slow SPR stored in 'ctrl'. Taken
branches and rfid update it to the address of the branch or rfid
instruction.
To simplify the logic, this makes rfid use the branch logic to
generate its redirect (requiring SRR0 to come in to execute1 on
the B input and SRR1 on the A input), and the masking of the bottom
2 bits of NIA is moved to fetch1.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Michael Neuling [Mon, 29 Jun 2020 02:19:06 +0000 (12:19 +1000)]
Merge pull request #213 from ozbenh/uart16550
Add support for standard 16550 style UART
Michael Neuling [Mon, 29 Jun 2020 02:18:44 +0000 (12:18 +1000)]
Merge pull request #212 from ozbenh/liteeth
liteeth: Hook up LiteX LiteEth ethernet controller
Michael Neuling [Mon, 29 Jun 2020 00:24:49 +0000 (10:24 +1000)]
Merge pull request #214 from shingarov/fix-ld-target
Fix ld error in elf maketarget
Boris Shingarov [Thu, 25 Jun 2020 09:31:45 +0000 (05:31 -0400)]
Fix ld error in elf maketarget
The sdram_init ELF fails to link:
powerpc64le-linux-gnu-ld -static -nostdlib -T sdram_init.lds \
--gc-sections -o sdram_init.elf head.o main.o sdram.o console.o \
libc.o sdram_init.lds
powerpc64le-linux-gnu-ld: error: linker script file 'sdram_init.lds'
appears multiple times
make: *** [Makefile:70: sdram_init.elf] Error 1
This is because sdram_init.lds is one of the prerequisites, and thus is
contained in $^. However, it is also explicitly specified as part of
LDFLAGS, as the argument to -T.
Signed-off-by: Boris Shingarov <shingarov@labware.com>
Benjamin Herrenschmidt [Fri, 19 Jun 2020 11:18:33 +0000 (21:18 +1000)]
tests: Add updated micropython build with 16550 support
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Fri, 19 Jun 2020 10:27:31 +0000 (20:27 +1000)]
sim_console: Fix polling to check for POLLIN
Under some circumstances we get POLLHUP which we incorrectly
treat as having a character in the buffer.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 18 Jun 2020 07:14:55 +0000 (17:14 +1000)]
uart: Make 16550 the default
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Tue, 23 Jun 2020 05:44:37 +0000 (15:44 +1000)]
syscon: Add flag to indicate the timebase frequency
This adds a flag (currently not set) to indicate that the core is using
the architected timebase frequency of 512Mhz. When not set, the core is
using the proc frequency for the timebase.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 18 Jun 2020 07:14:41 +0000 (17:14 +1000)]
console: Add support for the 16550 UART
And rebuild various binaries
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 18 Jun 2020 04:00:28 +0000 (14:00 +1000)]
uart: Add a simulation model for the 16550 compatible UART
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 18 Jun 2020 01:18:30 +0000 (11:18 +1000)]
uart: Rename sim_uart.vhdl to sim_pp_uart.vhdl
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 18 Jun 2020 01:06:33 +0000 (11:06 +1000)]
console: Cleanup console API
Use a more generic console_init() instead of potato_uart_init(),
and do the same for interrupt control. There should be no
change in behaviour.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Tue, 16 Jun 2020 12:42:15 +0000 (22:42 +1000)]
uart: Import and hook up opencore 16550 compatible UART
This imports via fusesoc a 16550 compatible (ie "standard") UART,
and wires it up optionally in the SoC instead of the potato one.
This also adds support for a second UART (which is always a
16550) to Arty, wired to JC "bottom" port.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Thu, 18 Jun 2020 09:41:00 +0000 (19:41 +1000)]
ex1: Add SPR_TBU support
It's used by the boot wrapper in Linux and possibly some userspace
programs.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 00:04:31 +0000 (10:04 +1000)]
liteeth: Hook up LiteX LiteEth ethernet controller
Currently only generated for Arty.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Neuling [Tue, 23 Jun 2020 06:58:06 +0000 (16:58 +1000)]
Merge pull request #211 from shenki/spi-constraint
spi: Fix dat_i_l constraints
Joel Stanley [Mon, 22 Jun 2020 09:03:14 +0000 (18:33 +0930)]
spi: Fix dat_i_l constraints
No cells matched 'get_cells -hierarchical -filter {NAME =~*/spi_rxtx/dat_i_l*}'. [build/microwatt_0/src/microwatt_0/fpga/arty_a7.xdc:42]
The signal is in it's own process so the net name ends up being
spi_rxtx/input_delay_1.dat_i_l_reg.
After this change the log shows:
Applied set_property IOB = TRUE for soc0/\spiflash_gen.spiflash /spi_rxtx/\input_delay_1.dat_i_l_reg . (constraint file fpga/arty_a7.xdc, line 42).
Applied set_property IOB = TRUE for soc0/\spiflash_gen.spiflash /spi_rxtx/\input_delay_1.dat_i_l_reg . (constraint file fpga/arty_a7.xdc, line 42).
Applied set_property IOB = TRUE for soc0/\spiflash_gen.spiflash /spi_rxtx/\input_delay_1.dat_i_l_reg . (constraint file fpga/arty_a7.xdc, line 42).
Applied set_property IOB = TRUE for soc0/\spiflash_gen.spiflash /spi_rxtx/\input_delay_1.dat_i_l_reg . (constraint file fpga/arty_a7.xdc, line 42).
Signed-off-by: Joel Stanley <joel@jms.id.au>
Michael Neuling [Tue, 23 Jun 2020 04:32:42 +0000 (14:32 +1000)]
Merge pull request #210 from ozbenh/xics
xics: Cleanups and add a simple ICS for use by Linux
Benjamin Herrenschmidt [Mon, 22 Jun 2020 13:38:34 +0000 (23:38 +1000)]
xics: Add support for reduced priority field size
This makes the ICS support less than the 8 architected bits
and sets the soc to use 3 bits by default.
All the supported bits set translates to "masked" (and will read
back at 0xff), any small value is used as-is.
Linux doesn't use priorities above 5, so this is a way to save
silicon. The number of supported priority bits is exposed to the
OS via the config register.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Wed, 17 Jun 2020 12:11:58 +0000 (22:11 +1000)]
xics: Add simple ICS
Move the external interrupt generation to a separate module
"ICS" (source controller) which a register per source containing
currently only the priority control.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Wed, 17 Jun 2020 12:07:33 +0000 (22:07 +1000)]
xics/icp: MFRR starts at 0xff not 0x00
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Wed, 17 Jun 2020 12:06:28 +0000 (22:06 +1000)]
tests/xics: Ensure no compiler optimisations in delay()
In case it would be tempted to "read ahead" the delay function
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Wed, 17 Jun 2020 11:51:16 +0000 (21:51 +1000)]
xics: ICP should be big endian !
That's how Linux expects it. This also simplifies the
register access implementation since the bit fields now
align properly regardless of the access size.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Wed, 17 Jun 2020 04:00:04 +0000 (14:00 +1000)]
tests: Fix Makefile.test to not allow host includes
xics was including the host limits.h for example
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Michael Neuling [Fri, 19 Jun 2020 01:50:47 +0000 (11:50 +1000)]
Merge pull request #208 from paulusmack/faster
Make the core go faster
Several major improvements in here:
- Simple branch predictor
- Reduced latency for mispredicted branches and interrupts by removing fetch2 stage
- Cache improvements
o Request critical dword first on refill
o Handle hits while refilling, including on line being refilled
o Sizes doubled for both D and I
- Loadstore improvements: can now do one load or store every two cycles in most cases
- Optimized 2-cycle multiplier for Xilinx 7-series parts using DSP slices
- Timing improvements, including:
o Stash buffer in decode1
o Reduced width of execute1 result mux
o Improved SPR decode in decode1
o Some non-critical operation take a cycle longer so we can break some long combinatorial chains
- Core logging: logs 256 bits of info every cycle into a ring buffer, to help with debugging and performance analysis
This increases the LUT usage for the "synth" + A35 target from 9182 to 10297 = 12%.
Paul Mackerras [Wed, 17 Jun 2020 21:26:01 +0000 (07:26 +1000)]
Merge pull request #207 from ozbenh/misc
Random cleanups of the SoC interfaces
Paul Mackerras [Tue, 16 Jun 2020 06:59:54 +0000 (16:59 +1000)]
fpga: Add a xilinx_specific fileset to microwatt.core
At present this just has the Xilinx-specific multiplier code, but
might in future have other things.
This also adds the xilinx_specific fileset to the synth target.
Without that it was failing because there was no multiplier.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Tue, 16 Jun 2020 01:37:25 +0000 (11:37 +1000)]
Make LOG_LENGTH configurable per FPGA variant
This plumbs the LOG_LENGTH parameter (which controls how many entries
the core log RAM has) up to the top level so that it can be set on
the fusesoc command line and have different default values on
different FPGAs.
It now defaults to 512 entries generally and on the Artix-7 35 parts,
and 2048 on the larger Artix-7 FPGAs. It can be set to 0 if desired.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Mon, 15 Jun 2020 06:59:08 +0000 (16:59 +1000)]
execute1: Reduce width of the result mux to help timing
This reduces the number of different things that are assigned to
the result variable.
- The computations for the popcnt, prty, cmpb and exts instruction
families are moved into the logical unit.
- The result of mfspr from the slow SPRs is computed in 'spr_val'
before being assigned to 'result'.
- Writes to LR as a result of a blr or bclr instruction are done
through the exc_write path to writeback.
This eases timing considerably.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Mon, 15 Jun 2020 05:43:05 +0000 (15:43 +1000)]
core: Implement a simple branch predictor
This implements a simple branch predictor in the decode1 stage. If it
sees that the instruction is b or bc and the branch is predicted to be
taken, it sends a flush and redirect upstream (to icache and fetch1)
to redirect fetching to the branch target. The prediction is sent
downstream with the branch instruction, and execute1 now only sends
a flush/redirect upstream if the prediction was wrong. Unconditional
branches are always predicted to be taken, and conditional branches
are predicted to be taken if and only if the offset is negative.
Branches that take the branch address from a register (bclr, bcctr)
are predicted not taken, as we don't have any way to predict the
branch address.
Since we can now have a mflr being executed immediately after a bl
or bcl, we now track the update to LR in the hazard tracker, using
the second write register field that is used to track RA updates for
update-form loads and stores.
For those branches that update LR but don't write any other result
(i.e. that don't decrementer CTR), we now write back LR in the same
cycle as the instruction rather than taking a second cycle for the
LR writeback.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Mon, 15 Jun 2020 00:02:14 +0000 (10:02 +1000)]
decode1: Improve timing for slow SPR decode path
This makes the logic that works out decode.unit and decode.sgl_pipe
for mtspr/mfspr to/from slow SPRs detect the fact that the
instruction is mtspr/mfspr based on a match with the instruction
word rather than looking at v.decode.insn_type. This improves timing
substantially, as the ROM lookup to get v.decode is relatively slow.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sun, 14 Jun 2020 23:28:03 +0000 (09:28 +1000)]
decode1: Add a stash buffer to the output
This means that the busy signal from execute1 (which can be driven
combinatorially from mmu or dcache) now stops at decode1 and doesn't
go on to icache or fetch1. This helps with timing.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 12:29:54 +0000 (22:29 +1000)]
soc: Slight cleanup of IRQ assignments
Use a separate process to assign selected interrupts to the
interrupt array, and document them.
There's only one interrupt *for now* but that will change
and this way is clearer.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 12:05:39 +0000 (22:05 +1000)]
soc: Rename uart_dat8 to uart0_dat8
Just for consistency. Will come in handy if we ever add a second one
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 12:04:45 +0000 (22:04 +1000)]
soc: Rename wb_dram_ctrl to wb_ext_io and rework decoding
This makes the control bus currently going out of "soc" towards
litedram more generic for external IO devices added by the
top-level rather than inside the SoC proper.
This is mostly renaming of signals and a small change on how the
address decoder operates, using a separate "cascaded" decode for
the external IOs.
We make the region 0xc8nn_nnnn be the "external IO" region for
now.
This will make it easier / cleaner to add more external devices.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Jonathan Balkind [Sat, 13 Jun 2020 22:39:18 +0000 (18:39 -0400)]
Minor refactor of icache to make less dependent on wishbone
Signed-off-by: Jonathan Balkind <jbalkind@princeton.edu>
Paul Mackerras [Sat, 13 Jun 2020 13:00:13 +0000 (23:00 +1000)]
dcache: Reduce back-to-back store latency from 3 cycles to 2
This uses the machinery we already had for comparing the real address
of a new request with the tag of a previous request (r1.reload_tag)
to get better timing on comparing the address of a second store with
the one in progress. The comparison is now on the set size rather
than the page size, but since set size can't be larger than the page
size (and usually will equal the page size), that is OK.
The same comparison can also be used to tell when we can satisfy
a load miss during a cache line refill.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 12:19:33 +0000 (22:19 +1000)]
soc: Don't require dram wishbones signals to be wired by toplevel
Currently, when not using litedram, the top level still has to hook
up "dummy" wishbones to the main dram and control dram busses coming
out of the SoC and provide ack signals.
Instead, make the SoC generate the acks internally when not using
litedram and use defaults to make the wiring entirely optional.
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 11:57:01 +0000 (21:57 +1000)]
soc: Add defaults for some input signals
That way the top-level's don't need to assign them
Also remove generics that are set to the default anyways
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Benjamin Herrenschmidt [Sat, 13 Jun 2020 11:51:31 +0000 (21:51 +1000)]
soc: Remove unused RESET_LOW generic
Signed-off-by: Benjamin Herrenschmidt <benh@kernel.crashing.org>
Paul Mackerras [Sat, 13 Jun 2020 10:27:50 +0000 (20:27 +1000)]
mmu: Take an extra cycle to do TLB invalidations
This makes the TLB invalidations that occur as a result of a tlbie,
slbia or mtspr instruction take one more cycle. This breaks some
long combinatorial chains from decode2 to dcache and icache and
thus eases timing.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 11 Jun 2020 04:23:50 +0000 (14:23 +1000)]
dcache: Reduce latencies and improve timing
This implements various improvements to the dcache with the aim of
making it go faster.
- We can now execute operations that don't need to access main memory
(cacheable loads that hit in the cache and TLB operations) as soon
as any previous operation has completed, without waiting for the
state machine to become idle.
- Cache line refills start with the doubleword that is needed to
satisfy the load that initiated them.
- Cacheable loads that miss return their data and complete as soon as
the requested doubleword comes back from memory; they don't wait for
the refill to finish.
- We now have per-doubleword valid bits for the cache line being
refilled, meaning that if a load comes in for a line that is in the
process of being refilled, we can return the data and complete it
within a couple of cycles of the doubleword coming in from memory.
- There is now a bypass path for data being written to the cache RAM
so that we can do a store hit followed immediately by a load hit to
the same doubleword. This also makes the data from a refill
available to load hits one cycle earlier than it would be otherwise.
- Stores complete in the cycle where their wishbone operation is
initiated, without waiting for the wishbone cycle to complete.
- During the wishbone cycle for a store, if another store comes in
that is to the same page, and we don't have a stall from the
wishbone, we can send out the write for the second store in the same
wishbone cycle and without going through the IDLE state first. We
limit it to 7 outstanding writes that have not yet been
acknowledged.
- The cache tag RAM is now read on a clock edge rather than being
combinatorial for reading. Its width is rounded up to a multiple of
8 bits per way so that byte enables can be used for writing
individual tags.
- The cache tag RAM is now written a cycle later than previously, in
order to ease timing.
- Data for a store hit is now written one cycle later than
previously. This eases timing since we don't have to get through
the tag matching and on to the write enable within a single cycle.
The 2-stage bypass path means we can still handle a load hit on
either of the two cycles after the store and return the correct
data. (A load hit 3 or more cycles later will get the correct data
from the BRAM.)
- Operations can sit in r0 while there is an uncompleted operation in
r1. Once the operation in r1 is completed, the operation in r0
spends one cycle in r0 for TLB/cache tag lookup and then gets put
into r1.req. This can happen before r1 gets to the IDLE state.
Some operations can then be completed before r1 gets to the IDLE
state - a load miss to the cache line being refilled, or a store to
the same page as a previous store.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Tue, 12 May 2020 06:28:42 +0000 (16:28 +1000)]
decode: Work out ispr1/ispr2 in parallel with decode ROM lookup
This makes the logic that calculates which SPRs are being accessed
work in parallel with the instruction decode ROM lookup instead of
being dependent on the opcode found in the decode ROM. The reason
for doing that is that the path from icache through the decode ROM
to the ispr1/ispr2 fields has become a critical path.
Thus we are now using only a very partial decode of the instruction
word in the logic for isp1/isp2, and we therefore can no longer rely
on them being zero in all cases where no SPR is being accessed.
Instead, decode2 now ignores ispr1/ispr2 in all cases except when the
relevant decode.input_reg_a/b or decode.output_reg_a is set to SPR.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 5 Jun 2020 04:22:02 +0000 (14:22 +1000)]
loadstore1: Reduce busy cycles
This reduces the number of cycles where loadstore1 asserts its busy
output, leading to increased throughput of loads and stores. Loads
that hit in the cache can now be executed at the rate of one every two
cycles. Stores take 4 cycles assuming the wishbone slave responds
with an ack the cycle after we assert strobe.
To achieve this, the state machine code is split into two parts, one
for when we have an existing instruction in progress, and one for
starting a new instruction. We can now combinatorially clear busy and
start a new instruction in the same cycle that we get a done signal
from the dcache; in other words we are completing one instruction and
potentially writing back results in the same cycle that we start a new
instruction and send its address and data to the dcache.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Fri, 5 Jun 2020 03:29:34 +0000 (13:29 +1000)]
loadstore1: Complete mfspr/mtspr a cycle later
This makes mfspr and mtspr complete (and mfspr write back) on the
cycle after the instruction is received from execute1, rather than
on the same cycle. This makes them match all other instructions
that execute in one cycle. Because these instructions are marked
as single-issue, there wasn't the possibility of having two
instructions complete on the same cycle (which we can't cope with),
but it is better to fix this.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 4 Jun 2020 10:58:32 +0000 (20:58 +1000)]
core: Use a busy signal rather than a stall
This changes the instruction dependency tracking so that we can
generate a "busy" signal from execute1 and loadstore1 which comes
along one cycle later than the current "stall" signal. This will
enable us to signal busy cycles only when we need to from loadstore1.
The "busy" signal from execute1/loadstore1 indicates "I didn't take
the thing you gave me on this cycle", as distinct from the previous
stall signal which meant "I took that but don't give me anything
next cycle". That means that decode2 proactively gives execute1
a new instruction as soon as it has taken the previous one (assuming
there is a valid instruction available from decode1), and that then
sits in decode2's output until execute1 can take it. So instructions
are issued by decode2 somewhat earlier than they used to be.
Decode2 now only signals a stall upstream when its output buffer is
full, meaning that we can fill up bubbles in the upstream pipe while a
long instruction is executing. This gives a small boost in
performance.
This also adds dependency tracking for rA updates by update-form
load/store instructions.
The GPR and CR hazard detection machinery now has one extra stage,
which may not be strictly necessary. Some of the code now really
only applies to PIPELINE_DEPTH=1.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 28 May 2020 23:38:05 +0000 (09:38 +1000)]
icache: Improve latencies when reloading cache lines
The icache can now detect a hit on a line being refilled from memory,
as we have an array of individual valid bits per row for the line
that is currently being loaded. This enables the request that
initiated the refill to be satisfied earlier, and also enables
following requests to the same cache line to be satisfied before the
line is completely refilled. Furthermore, the refill now starts
at the row that is needed. This should reduce the latency for an
icache miss.
We now get a 'sequential' indication from fetch1, and use that to know
when we can deliver an instruction word using the other half of the
64-bit doubleword that was read last cycle. This doesn't make much
difference at the moment, but it frees up cycles where we could test
whether the next line is present in the cache so that we could
prefetch it if not.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 21 May 2020 07:50:54 +0000 (17:50 +1000)]
multiply: Use DSP48 slices for multiplication on Xilinx FPGAs
This adds a custom implementation of the multiplier which uses 16
DSP48E1 slices to do a 64x64 bit multiplication in 2 cycles.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 21 May 2020 03:42:46 +0000 (13:42 +1000)]
multiply: Move selection of result bits into execute1
This puts the logic that selects which bits of the multiplier result
get written into the destination GPR into execute1, moved out from
multiply.
The multiplier is now expected to do an unsigned multiplication of
64-bit operands, optionally negate the result, detect 32-bit
or 64-bit signed overflow of the result, and return a full 128-bit
result.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Wed, 3 Jun 2020 01:26:33 +0000 (11:26 +1000)]
core: Double the dcache and icache sizes
This makes the dcache and icache both be 8kB. This still only uses
one BRAM per way per cache on the Artix-7, since the BRAMs were only
half-used previously.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sun, 10 May 2020 08:18:03 +0000 (18:18 +1000)]
core: Remove fetch2 pipeline stage
The fetch2 stage existed primarily to provide a stash buffer for the
output of icache when a stall occurred. However, we can get the same
effect -- of having the input to decode1 stay unchanged on a stall
cycle -- by using the read enable of the BRAMs in icache, and by
adding logic to keep the outputs unchanged on a clock cycle when
stall_in = 1. This reduces branch and interrupt latency by one
cycle.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 14 May 2020 03:25:48 +0000 (13:25 +1000)]
Add core logging
This logs 256 bits of data per cycle to a ring buffer in BRAM. The
data collected can be read out through 2 new SPRs or through the
debug interface.
The new SPRs are LOG_ADDR (724) and LOG_DATA (725). LOG_ADDR contains
the buffer write pointer in the upper 32 bits (in units of entries,
i.e. 32 bytes) and the read pointer in the lower 32 bits (in units of
doublewords, i.e. 8 bytes). Reading LOG_DATA gives the doubleword
from the buffer at the read pointer and increments the read pointer.
Setting bit 31 of LOG_ADDR inhibits the trace log system from writing
to the log buffer, so the contents are stable and can be read.
There are two new debug addresses which function similarly to the
LOG_ADDR and LOG_DATA SPRs. The log is frozen while either or both of
the LOG_ADDR SPR bit 31 or the debug LOG_ADDR register bit 31 are set.
The buffer defaults to 2048 entries, i.e. 64kB. The size is set by
the LOG_LENGTH generic on the core_debug module. Software can
determine the length of the buffer because the length is ORed into the
buffer write pointer in the upper 32 bits of LOG_ADDR. Hence the
length of the buffer can be calculated as 1 << (31 - clz(LOG_ADDR)).
There is a program to format the log entries in a somewhat readable
fashion in scripts/fmt_log/fmt_log.c. The log_entry struct in that
file describes the layout of the bits in the log entries.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 13 Jun 2020 09:59:17 +0000 (19:59 +1000)]
scripts/mw_debug: Make progress counts display on one line
This outputs a carriage return rather than a newline after the
display of the progress count during the load and save operations.
This makes the output more compact and better looking.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Thu, 11 Jun 2020 04:57:30 +0000 (14:57 +1000)]
tests/xics: Fix assumption that interrupts happen immediately
Currently the test writes to the XICS and then checks that the
expected interrupt has happened. This turns into a stbcix
instruction followed immediately by a load from the variable that
indicates whether an interrupt has happened. It is possible for
it to take a few cycles for the store to reach the XICS and the
interrupt request signal to come back to the core, particularly
with improvements to the load/store unit and dcache.
This therefore adds a delay between storing to the XICS and
checking for the occurrence of an interrupt, so as to give the
signals time to propagate. The delay loop does an arbitrary 10
iterations, and each iteration does two loads and one store to
(cacheable) memory.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 13 Jun 2020 07:22:56 +0000 (17:22 +1000)]
register_file: Report value being written before asserting it's not X
If a bug causes an indeterminate value to be written to a GPR, an
assert causes simulation to abort. Move the assert after the report
of the GPR index and value so that we get to know what the bad value
is before the simulation terminates.
Signed-off-by: Paul Mackerras <paulus@ozlabs.org>
Paul Mackerras [Sat, 13 Jun 2020 02:36:16 +0000 (12:36 +1000)]
Merge pull request #205 from ozbenh/timing
Timing improvements