SeaBIOS May 2014

seabios@seabios.org

16 participants
46 discussions

Re: [SeaBIOS] [Xen-devel] [PATCH v3] Hvmloader: Modify ACPI to only supply _EJ0 methods for PCIslots that support hotplug by runtime patching

by Johannes Krampf

>> > Well, Xen code is under GPLv2 so I don't think you can do this? >> > >> > Unless Michael is willing to release it under a different licence? >> >> I'd be fine with "GPLv2 or later" so code can flow back to seabios >> as well - but you will have to track down everyone else who >> changed this file since I wrote it and get their permission. >> > Thanks, Michael. I search the git log of above two python files, and > find two person has changed those two files, Johannes and Kevin. > > CC'ing Johannes and Kevin for their permission, which using those > two files in Xen code. Thanks in advance! I hereby grant you permission to use the code changes I contributed to SeaBIOS under a "GPLv2 and later" license. - Johannes

6 years, 5 months

Re: [SeaBIOS] [Xen-devel] [PATCH v3] Hvmloader: Modify ACPI to only supply _EJ0 methods for PCIslots that support hotplug by runtime patching

by Gonglei (Arei)

Hi, all > -----Original Message----- > From: Michael S. Tsirkin [mailto:mst@redhat.com] > Sent: Thursday, May 08, 2014 4:15 AM > To: Konrad Rzeszutek Wilk > Cc: Gonglei (Arei); qemu-devel(a)nongnu.org; xen-devel(a)lists.xen.org; > Huangweidong (C); Ian.Campbell(a)citrix.com; stefano.stabellini(a)eu.citrix.com; > Hanweidong (Randy); fabio.fantoni(a)m2r.biz; JBeulich(a)suse.com; > anthony.perard(a)citrix.com; Gaowei (UVP) > Subject: Re: [Xen-devel] [PATCH v3] Hvmloader: Modify ACPI to only supply _EJ0 > methods for PCIslots that support hotplug by runtime patching > > On Wed, May 07, 2014 at 02:48:48PM -0400, Konrad Rzeszutek Wilk wrote: > > On Sun, May 04, 2014 at 05:25:15PM +0800, arei.gonglei(a)huawei.com > wrote: > > > From: Gaowei <gao.gaowei(a)huawei.com> > > > > > > In Xen platform, after using upstream qemu, the all of pci devices will show > > > hotplug in the windows guest. In this situation, the windows guest may > occur > > > blue screen when VM' user click the icon of VGA card for trying unplug VGA > card. > > > However, we don't hope VM's user can do such dangerous operation, and > showing > > > all pci devices inside the guest OS is unfriendly. > > > > > > This is done by runtime patching: > > > - Rename _EJ0 methods for PCI slots in DSDT to EJ0_:note that this has > the > > > same checksum, but is ignored by OSPM. > > > - At compile time, look for these methods in ASL source,find the > matching AML, > > > and store the offsets of these methods in a table named > aml_ej0_data. > > > - At run time, go over aml_ej0_data, check which slots not support > hotplug and > > > patch the ACPI table, replacing _EJ0 with EJ0_. > > > > How does this work with traditional QEMU ? Is this only when running with > SeaBIOS and upstream QEMU? > > The traditional QEMU is unaffected and works fine as normal. Yes, it is only useful With seabios and upstream QEMU. > > > > > > Signed-off-by: Gaowei <gao.gaowei(a)huawei.com> > > > Signed-off-by: Gonglei <arei.gonglei(a)huawei.com> > > > --- > > > v3->v2: > > > reformat on the latest master > > > v2->v1: > > > rework by Jan Beulich's suggestion: > > > - adjust the code style > > > > > > tools/firmware/hvmloader/acpi/Makefile | 44 ++- > > > tools/firmware/hvmloader/acpi/acpi2_0.h | 4 + > > > tools/firmware/hvmloader/acpi/build.c | 22 ++ > > > tools/firmware/hvmloader/acpi/dsdt.asl | 1 + > > > tools/firmware/hvmloader/acpi/mk_dsdt.c | 2 + > > > tools/firmware/hvmloader/ovmf.c | 6 +- > > > tools/firmware/hvmloader/rombios.c | 4 + > > > tools/firmware/hvmloader/seabios.c | 8 + > > > tools/firmware/hvmloader/tools/acpi_extract.py | 308 > +++++++++++++++++++++ > > > .../hvmloader/tools/acpi_extract_preprocess.py | 41 +++ > > > 10 files changed, 428 insertions(+), 12 deletions(-) > > > create mode 100644 tools/firmware/hvmloader/tools/acpi_extract.py > > > create mode 100644 > tools/firmware/hvmloader/tools/acpi_extract_preprocess.py > > > > > > diff --git a/tools/firmware/hvmloader/acpi/Makefile > b/tools/firmware/hvmloader/acpi/Makefile > > > index 2c50851..f79ecc3 100644 > > > --- a/tools/firmware/hvmloader/acpi/Makefile > > > +++ b/tools/firmware/hvmloader/acpi/Makefile > > > @@ -24,30 +24,46 @@ OBJS = $(patsubst %.c,%.o,$(C_SRC)) > > > CFLAGS += $(CFLAGS_xeninclude) > > > > > > vpath iasl $(PATH) > > > +vpath python $(PATH) > > > + > > > +.DELETE_ON_ERROR:$(filter dsdt_%.c,$(C_SRC)) > > > + > > > all: acpi.a > > > > > > ssdt_s3.h ssdt_s4.h ssdt_pm.h ssdt_tpm.h: %.h: %.asl iasl > > > iasl -vs -p $* -tc $< > > > - sed -e 's/AmlCode/$*/g' $*.hex >$@ > > > + sed -e 's/AmlCode/$*/g' $*.hex >$@.tmp > > > + $(call move-if-changed,$@.tmp $@) > > > rm -f $*.hex $*.aml > > > > > > mk_dsdt: mk_dsdt.c > > > $(HOSTCC) $(HOSTCFLAGS) $(CFLAGS_xeninclude) -o $@ mk_dsdt.c > > > > > > dsdt_anycpu_qemu_xen.asl: dsdt.asl mk_dsdt > > > - awk 'NR > 1 {print s} {s=$$0}' $< > $@ > > > - ./mk_dsdt --dm-version qemu-xen >> $@ > > > + awk 'NR > 1 {print s} {s=$$0}' $< > $@.tmp > > > + sed -i 's/AmlCode/dsdt_anycpu_qemu_xen/g' $@.tmp > > > + ./mk_dsdt --dm-version qemu-xen >> $@.tmp > > > + sed -i 's/aml_ej0_name/dsdt_anycpu_qemu_xen_aml_ej0_name/g' > $@.tmp > > > + sed -i 's/aml_adr_dword/dsdt_anycpu_qemu_xen_aml_adr_dword/g' > $@.tmp > > > + $(call move-if-changed,$@.tmp $@) > > > > > > # NB. awk invocation is a portable alternative to 'head -n -1' > > > dsdt_%cpu.asl: dsdt.asl mk_dsdt > > > - awk 'NR > 1 {print s} {s=$$0}' $< > $@ > > > - ./mk_dsdt --maxcpu $* >> $@ > > > + awk 'NR > 1 {print s} {s=$$0}' $< > $@.tmp > > > + sed -i 's/AmlCode/dsdt_$*cpu/g' $@.tmp > > > + ./mk_dsdt --maxcpu $* >> $@.tmp > > > + $(call move-if-changed,$@.tmp $@) > > > > > > -$(filter dsdt_%.c,$(C_SRC)): %.c: iasl %.asl > > > - iasl -vs -p $* -tc $*.asl > > > - sed -e 's/AmlCode/$*/g' $*.hex >$@ > > > - echo "int $*_len=sizeof($*);" >>$@ > > > - rm -f $*.aml $*.hex > > > +$(filter dsdt_%.c,$(C_SRC)): %.c: iasl python %.asl > > > + cpp -P $*.asl > $*.asl.i.orig > > > + python ../tools/acpi_extract_preprocess.py $*.asl.i.orig > $*.asl.i > > > + iasl -vs -l -tc -p $* $*.asl.i > > > + python ../tools/acpi_extract.py $*.lst > $@.tmp > > > + echo "int $*_len=sizeof($*);" >>$@.tmp > > > + if grep -q "$*_aml_ej0_name" $@.tmp; then echo "int > $*_aml_ej0_name_len=sizeof($*_aml_ej0_name);" >>$@.tmp; fi > > > + if grep -q "$*_aml_adr_dword" $@.tmp; then echo "int > $*_aml_adr_dword_len=sizeof($*_aml_adr_dword);" >>$@.tmp;fi > > > + $(call move-if-changed,$@.tmp $@) > > > + rm -f $*.aml $*.hex $*.asl.i.orig $*.asl.i $*.lst > > > > > > iasl: > > > @echo > > > @@ -57,6 +73,12 @@ iasl: > > > @echo > > > @exit 1 > > > > > > +python: > > > + @echo > > > + @echo "python is needed" > > > + @echo > > > + @exit 1 > > > + > > > build.o: ssdt_s3.h ssdt_s4.h ssdt_pm.h ssdt_tpm.h > > > > > > acpi.a: $(OBJS) > > > @@ -64,7 +86,7 @@ acpi.a: $(OBJS) > > > > > > clean: > > > rm -rf *.a *.o $(IASL_VER) $(IASL_VER).tar.gz $(DEPS) > > > - rm -rf ssdt_*.h dsdt*.c *~ *.aml *.hex mk_dsdt dsdt_*.asl > > > + rm -rf ssdt_*.h dsdt*.c *~ *.aml *.hex mk_dsdt dsdt_*.asl *.asl.i > *.asl.i.orig *.lst *.tmp > > > > > > install: all > > > > > > diff --git a/tools/firmware/hvmloader/acpi/acpi2_0.h > b/tools/firmware/hvmloader/acpi/acpi2_0.h > > > index 7b22d80..4ba3957 100644 > > > --- a/tools/firmware/hvmloader/acpi/acpi2_0.h > > > +++ b/tools/firmware/hvmloader/acpi/acpi2_0.h > > > @@ -396,6 +396,10 @@ struct acpi_config { > > > int dsdt_anycpu_len; > > > unsigned char *dsdt_15cpu; > > > int dsdt_15cpu_len; > > > + unsigned short *aml_ej0_name; > > > + unsigned short *aml_adr_dword; > > > + int aml_ej0_name_len; > > > + int aml_adr_dword_len; > > > }; > > > > > > void acpi_build_tables(struct acpi_config *config, unsigned int physical); > > > diff --git a/tools/firmware/hvmloader/acpi/build.c > b/tools/firmware/hvmloader/acpi/build.c > > > index f1dd3f0..ccce420 100644 > > > --- a/tools/firmware/hvmloader/acpi/build.c > > > +++ b/tools/firmware/hvmloader/acpi/build.c > > > @@ -30,6 +30,8 @@ > > > #define align16(sz) (((sz) + 15) & ~15) > > > #define fixed_strcpy(d, s) strncpy((d), (s), sizeof(d)) > > > > > > +#define PCI_RMV_BASE 0xae0c > > > + > > > extern struct acpi_20_rsdp Rsdp; > > > extern struct acpi_20_rsdt Rsdt; > > > extern struct acpi_20_xsdt Xsdt; > > > @@ -404,6 +406,8 @@ void acpi_build_tables(struct acpi_config *config, > unsigned int physical) > > > unsigned char *dsdt; > > > unsigned long > secondary_tables[ACPI_MAX_SECONDARY_TABLES]; > > > int nr_secondaries, i; > > > + unsigned int rmvc_pcrm = 0; > > > + unsigned int len_aml_addr = 0, len_aml_ej0 = 0; > > > > > > /* Allocate and initialise the acpi info area. */ > > > mem_hole_populate_ram(ACPI_INFO_PHYSICAL_ADDRESS >> > PAGE_SHIFT, 1); > > > @@ -440,6 +444,24 @@ void acpi_build_tables(struct acpi_config *config, > unsigned int physical) > > > memcpy(dsdt, config->dsdt_anycpu, config->dsdt_anycpu_len); > > > nr_processor_objects = HVM_MAX_VCPUS; > > > } > > > + len_aml_addr = > config->aml_adr_dword_len/sizeof(config->aml_adr_dword[0]); > > > + len_aml_ej0 = > config->aml_ej0_name_len/sizeof(config->aml_ej0_name[0]); > > > + if (config->aml_adr_dword_len && config->aml_ej0_name_len && > (len_aml_addr == len_aml_ej0)) > > > + { > > > + rmvc_pcrm = inl(PCI_RMV_BASE); > > > > > > So .. what is at this special 0xae0c address? > > > > Is there some code in QEMU that needs this treatment as well? > > Yep, the 0xae0c is defined in upstream QEMU, which show the hotplugging ability of PCI devices. Including other special address about acpi hotplug defined in hw/acpi/pcihp.c: #define ACPI_PCI_HOTPLUG_STATUS 2 #define ACPI_PCIHP_ADDR 0xae00 #define ACPI_PCIHP_SIZE 0x0014 #define ACPI_PCIHP_LEGACY_SIZE 0x000f #define PCI_UP_BASE 0x0000 #define PCI_DOWN_BASE 0x0004 #define PCI_EJ_BASE 0x0008 #define PCI_RMV_BASE 0x000c #define PCI_SEL_BASE 0x0010 > > > + printf("rmvc_pcrm is %x\n", rmvc_pcrm); > > > + for (i = 0; i < len_aml_addr; i++ ) { > > > + /* Slot is in byte 2 in _ADR */ > > > + unsigned char slot = dsdt[config->aml_adr_dword[i] + 2] & > 0x1F; > > > + /* Sanity check */ > > > + if (memcmp(dsdt + config->aml_ej0_name[i], "_EJ0", 4)) { > > > + goto oom; > > > + } > > > + if (!(rmvc_pcrm & (0x1 << slot))) { > > > + memcpy(dsdt + config->aml_ej0_name[i], "EJ0_", 4); > > > + } > > > + } > > > + } > > > > > > /* > > > * N.B. ACPI 1.0 operating systems may not handle FADT with > revision 2 > > > diff --git a/tools/firmware/hvmloader/acpi/dsdt.asl > b/tools/firmware/hvmloader/acpi/dsdt.asl > > > index 247a8ad..1e7695b 100644 > > > --- a/tools/firmware/hvmloader/acpi/dsdt.asl > > > +++ b/tools/firmware/hvmloader/acpi/dsdt.asl > > > @@ -16,6 +16,7 @@ > > > * this program; if not, write to the Free Software Foundation, Inc., 59 > Temple > > > * Place - Suite 330, Boston, MA 02111-1307 USA. > > > */ > > > +ACPI_EXTRACT_ALL_CODE AmlCode > > > > > > DefinitionBlock ("DSDT.aml", "DSDT", 2, "Xen", "HVM", 0) > > > { > > > diff --git a/tools/firmware/hvmloader/acpi/mk_dsdt.c > b/tools/firmware/hvmloader/acpi/mk_dsdt.c > > > index a4b693b..555e062 100644 > > > --- a/tools/firmware/hvmloader/acpi/mk_dsdt.c > > > +++ b/tools/firmware/hvmloader/acpi/mk_dsdt.c > > > @@ -372,7 +372,9 @@ int main(int argc, char **argv) > > > /* hotplug_slot */ > > > for (slot = 1; slot <= 31; slot++) { > > > push_block("Device", "S%i", slot); { > > > + printf("ACPI_EXTRACT_NAME_DWORD_CONST > aml_adr_dword\n"); > > > stmt("Name", "_ADR, %#06x0000", slot); > > > + printf("ACPI_EXTRACT_METHOD_STRING > aml_ej0_name\n"); > > > push_block("Method", "_EJ0,1"); { > > > stmt("Store", "ShiftLeft(1, %#06x), B0EJ", slot); > > > stmt("Return", "0x0"); > > > diff --git a/tools/firmware/hvmloader/ovmf.c > b/tools/firmware/hvmloader/ovmf.c > > > index 28dd7bc..ea11564 100644 > > > --- a/tools/firmware/hvmloader/ovmf.c > > > +++ b/tools/firmware/hvmloader/ovmf.c > > > @@ -123,7 +123,11 @@ static void ovmf_acpi_build_tables(void) > > > .dsdt_anycpu = dsdt_anycpu, > > > .dsdt_anycpu_len = dsdt_anycpu_len, > > > .dsdt_15cpu = NULL, > > > - .dsdt_15cpu_len = 0 > > > + .dsdt_15cpu_len = 0, > > > + .aml_ej0_name = NULL, > > > + .aml_adr_dword = NULL, > > > + .aml_ej0_name_len = 0, > > > + .aml_adr_dword_len = 0, > > > }; > > > > > > acpi_build_tables(&config, ACPI_PHYSICAL_ADDRESS); > > > diff --git a/tools/firmware/hvmloader/rombios.c > b/tools/firmware/hvmloader/rombios.c > > > index 810bd24..803c9fa 100644 > > > --- a/tools/firmware/hvmloader/rombios.c > > > +++ b/tools/firmware/hvmloader/rombios.c > > > @@ -179,6 +179,10 @@ static void rombios_acpi_build_tables(void) > > > .dsdt_anycpu_len = dsdt_anycpu_len, > > > .dsdt_15cpu = dsdt_15cpu, > > > .dsdt_15cpu_len = dsdt_15cpu_len, > > > + .aml_ej0_name = NULL, > > > + .aml_adr_dword = NULL, > > > + .aml_ej0_name_len = 0, > > > + .aml_adr_dword_len = 0, > > > }; > > > > > > acpi_build_tables(&config, ACPI_PHYSICAL_ADDRESS); > > > diff --git a/tools/firmware/hvmloader/seabios.c > b/tools/firmware/hvmloader/seabios.c > > > index dd7dfbe..ca01d27 100644 > > > --- a/tools/firmware/hvmloader/seabios.c > > > +++ b/tools/firmware/hvmloader/seabios.c > > > @@ -33,6 +33,10 @@ > > > > > > extern unsigned char dsdt_anycpu_qemu_xen[]; > > > extern int dsdt_anycpu_qemu_xen_len; > > > +extern unsigned short dsdt_anycpu_qemu_xen_aml_ej0_name[]; > > > +extern unsigned short dsdt_anycpu_qemu_xen_aml_adr_dword[]; > > > +extern int dsdt_anycpu_qemu_xen_aml_ej0_name_len; > > > +extern int dsdt_anycpu_qemu_xen_aml_adr_dword_len; > > > > > > struct seabios_info { > > > char signature[14]; /* XenHVMSeaBIOS\0 */ > > > @@ -99,6 +103,10 @@ static void seabios_acpi_build_tables(void) > > > .dsdt_anycpu_len = dsdt_anycpu_qemu_xen_len, > > > .dsdt_15cpu = NULL, > > > .dsdt_15cpu_len = 0, > > > + .aml_ej0_name = dsdt_anycpu_qemu_xen_aml_ej0_name, > > > + .aml_adr_dword = dsdt_anycpu_qemu_xen_aml_adr_dword, > > > + .aml_ej0_name_len = > dsdt_anycpu_qemu_xen_aml_ej0_name_len, > > > + .aml_adr_dword_len = > dsdt_anycpu_qemu_xen_aml_adr_dword_len, > > > }; > > > > > > acpi_build_tables(&config, rsdp); > > > diff --git a/tools/firmware/hvmloader/tools/acpi_extract.py > b/tools/firmware/hvmloader/tools/acpi_extract.py > > > new file mode 100644 > > > index 0000000..ccec089 > > > --- /dev/null > > > +++ b/tools/firmware/hvmloader/tools/acpi_extract.py > > > @@ -0,0 +1,308 @@ > > > +#!/usr/bin/python > > > +# Copyright (C) 2011 Red Hat, Inc., Michael S. Tsirkin <mst(a)redhat.com> > > > +# > > > +# This file may be distributed under the terms of the GNU GPLv3 license. > > > > Well, Xen code is under GPLv2 so I don't think you can do this? > > > > Unless Michael is willing to release it under a different licence? > > I'd be fine with "GPLv2 or later" so code can flow back to seabios > as well - but you will have to track down everyone else who > changed this file since I wrote it and get their permission. > Thanks, Michael. I search the git log of above two python files, and find two person has changed those two files, Johannes and Kevin. CC'ing Johannes and Kevin for their permission, which using those two files in Xen code. Thanks in advance! CC'ing seabios maillist too. Best regards, -Gonglei > > > + > > > +# Process mixed ASL/AML listing (.lst file) produced by iasl -l > > > +# Locate and execute ACPI_EXTRACT directives, output offset info > > > +# > > > +# Documentation of ACPI_EXTRACT_* directive tags: > > > +# > > > +# These directive tags output offset information from AML for BIOS > runtime > > > +# table generation. > > > +# Each directive is of the form: > > > +# ACPI_EXTRACT_<TYPE> <array_name> <Operator> (...) > > > +# and causes the extractor to create an array > > > +# named <array_name> with offset, in the generated AML, > > > +# of an object of a given type in the following <Operator>. > > > +# > > > +# A directive must fit on a single code line. > > > +# > > > +# Object type in AML is verified, a mismatch causes a build failure. > > > +# > > > +# Directives and operators currently supported are: > > > +# ACPI_EXTRACT_NAME_DWORD_CONST - extract a Dword Const object > from Name() > > > +# ACPI_EXTRACT_NAME_WORD_CONST - extract a Word Const object > from Name() > > > +# ACPI_EXTRACT_NAME_BYTE_CONST - extract a Byte Const object from > Name() > > > +# ACPI_EXTRACT_METHOD_STRING - extract a NameString from > Method() > > > +# ACPI_EXTRACT_NAME_STRING - extract a NameString from Name() > > > +# ACPI_EXTRACT_PROCESSOR_START - start of Processor() block > > > +# ACPI_EXTRACT_PROCESSOR_STRING - extract a NameString from > Processor() > > > +# ACPI_EXTRACT_PROCESSOR_END - offset at last byte of Processor() + 1 > > > +# ACPI_EXTRACT_PKG_START - start of Package block > > > +# > > > +# ACPI_EXTRACT_ALL_CODE - create an array storing the generated AML > bytecode > > > +# > > > +# ACPI_EXTRACT is not allowed anywhere else in code, except in > comments. > > > + > > > +import re; > > > +import sys; > > > +import fileinput; > > > + > > > +aml = [] > > > +asl = [] > > > +output = {} > > > +debug = "" > > > + > > > +class asl_line: > > > + line = None > > > + lineno = None > > > + aml_offset = None > > > + > > > +def die(diag): > > > + sys.stderr.write("Error: %s; %s\n" % (diag, debug)) > > > + sys.exit(1) > > > + > > > +#Store an ASL command, matching AML offset, and input line (for > debugging) > > > +def add_asl(lineno, line): > > > + l = asl_line() > > > + l.line = line > > > + l.lineno = lineno > > > + l.aml_offset = len(aml) > > > + asl.append(l) > > > + > > > +#Store an AML byte sequence > > > +#Verify that offset output by iasl matches # of bytes so far > > > +def add_aml(offset, line): > > > + o = int(offset, 16); > > > + # Sanity check: offset must match size of code so far > > > + if (o != len(aml)): > > > + die("Offset 0x%x != 0x%x" % (o, len(aml))) > > > + # Strip any trailing dots and ASCII dump after " > > > + line = re.sub(r'\s*\.*\s*".*$',"", line) > > > + # Strip traling whitespace > > > + line = re.sub(r'\s+$',"", line) > > > + # Strip leading whitespace > > > + line = re.sub(r'^\s+',"", line) > > > + # Split on whitespace > > > + code = re.split(r'\s+', line) > > > + for c in code: > > > + # Require a legal hex number, two digits > > > + if (not(re.search(r'^[0-9A-Fa-f][0-9A-Fa-f]$', c))): > > > + die("Unexpected octet %s" % c); > > > + aml.append(int(c, 16)); > > > + > > > +# Process aml bytecode array, decoding AML > > > +def aml_pkglen_bytes(offset): > > > + # PkgLength can be multibyte. Bits 8-7 give the # of extra bytes. > > > + pkglenbytes = aml[offset] >> 6; > > > + return pkglenbytes + 1 > > > + > > > +def aml_pkglen(offset): > > > + pkgstart = offset > > > + pkglenbytes = aml_pkglen_bytes(offset) > > > + pkglen = aml[offset] & 0x3F > > > + # If multibyte, first nibble only uses bits 0-3 > > > + if ((pkglenbytes > 0) and (pkglen & 0x30)): > > > + die("PkgLen bytes 0x%x but first nibble 0x%x expected 0x0X" % > > > + (pkglen, pkglen)) > > > + offset += 1 > > > + pkglenbytes -= 1 > > > + for i in range(pkglenbytes): > > > + pkglen |= aml[offset + i] << (i * 8 + 4) > > > + if (len(aml) < pkgstart + pkglen): > > > + die("PckgLen 0x%x at offset 0x%x exceeds AML size 0x%x" % > > > + (pkglen, offset, len(aml))) > > > + return pkglen > > > + > > > +# Given method offset, find its NameString offset > > > +def aml_method_string(offset): > > > + #0x14 MethodOp PkgLength NameString MethodFlags TermList > > > + if (aml[offset] != 0x14): > > > + die( "Method offset 0x%x: expected 0x14 actual 0x%x" % > > > + (offset, aml[offset])); > > > + offset += 1; > > > + pkglenbytes = aml_pkglen_bytes(offset) > > > + offset += pkglenbytes; > > > + return offset; > > > + > > > +# Given name offset, find its NameString offset > > > +def aml_name_string(offset): > > > + #0x08 NameOp NameString DataRef > > > + if (aml[offset] != 0x08): > > > + die( "Name offset 0x%x: expected 0x08 actual 0x%x" % > > > + (offset, aml[offset])); > > > + offset += 1 > > > + # Block Name Modifier. Skip it. > > > + if (aml[offset] == 0x5c or aml[offset] == 0x5e): > > > + offset += 1 > > > + return offset; > > > + > > > +# Given data offset, find dword const offset > > > +def aml_data_dword_const(offset): > > > + #0x08 NameOp NameString DataRef > > > + if (aml[offset] != 0x0C): > > > + die( "Name offset 0x%x: expected 0x0C actual 0x%x" % > > > + (offset, aml[offset])); > > > + return offset + 1; > > > + > > > +# Given data offset, find word const offset > > > +def aml_data_word_const(offset): > > > + #0x08 NameOp NameString DataRef > > > + if (aml[offset] != 0x0B): > > > + die( "Name offset 0x%x: expected 0x0B actual 0x%x" % > > > + (offset, aml[offset])); > > > + return offset + 1; > > > + > > > +# Given data offset, find byte const offset > > > +def aml_data_byte_const(offset): > > > + #0x08 NameOp NameString DataRef > > > + if (aml[offset] != 0x0A): > > > + die( "Name offset 0x%x: expected 0x0A actual 0x%x" % > > > + (offset, aml[offset])); > > > + return offset + 1; > > > + > > > +# Given name offset, find dword const offset > > > +def aml_name_dword_const(offset): > > > + return aml_data_dword_const(aml_name_string(offset) + 4) > > > + > > > +# Given name offset, find word const offset > > > +def aml_name_word_const(offset): > > > + return aml_data_word_const(aml_name_string(offset) + 4) > > > + > > > +# Given name offset, find byte const offset > > > +def aml_name_byte_const(offset): > > > + return aml_data_byte_const(aml_name_string(offset) + 4) > > > + > > > +def aml_processor_start(offset): > > > + #0x5B 0x83 ProcessorOp PkgLength NameString ProcID > > > + if ((aml[offset] != 0x5B) or (aml[offset + 1] != 0x83)): > > > + die( "Name offset 0x%x: expected 0x5B 0x83 actual 0x%x > 0x%x" % > > > + (offset, aml[offset], aml[offset + 1])); > > > + return offset > > > + > > > +def aml_processor_string(offset): > > > + #0x5B 0x83 ProcessorOp PkgLength NameString ProcID > > > + start = aml_processor_start(offset) > > > + offset += 2 > > > + pkglenbytes = aml_pkglen_bytes(offset) > > > + offset += pkglenbytes > > > + return offset > > > + > > > +def aml_processor_end(offset): > > > + start = aml_processor_start(offset) > > > + offset += 2 > > > + pkglenbytes = aml_pkglen_bytes(offset) > > > + pkglen = aml_pkglen(offset) > > > + return offset + pkglen > > > + > > > +def aml_package_start(offset): > > > + offset = aml_name_string(offset) + 4 > > > + # 0x12 PkgLength NumElements PackageElementList > > > + if (aml[offset] != 0x12): > > > + die( "Name offset 0x%x: expected 0x12 actual 0x%x" % > > > + (offset, aml[offset])); > > > + offset += 1 > > > + return offset + aml_pkglen_bytes(offset) + 1 > > > + > > > +lineno = 0 > > > +for line in fileinput.input(): > > > + # Strip trailing newline > > > + line = line.rstrip(); > > > + # line number and debug string to output in case of errors > > > + lineno = lineno + 1 > > > + debug = "input line %d: %s" % (lineno, line) > > > + #ASL listing: space, then line#, then ...., then code > > > + pasl = re.compile('^\s+([0-9]+)\.\.\.\.\s*') > > > + m = pasl.search(line) > > > + if (m): > > > + add_asl(lineno, pasl.sub("", line)); > > > + # AML listing: offset in hex, then ...., then code > > > + paml = re.compile('^([0-9A-Fa-f]+)\.\.\.\.\s*') > > > + m = paml.search(line) > > > + if (m): > > > + add_aml(m.group(1), paml.sub("", line)) > > > + > > > +# Now go over code > > > +# Track AML offset of a previous non-empty ASL command > > > +prev_aml_offset = -1 > > > +for i in range(len(asl)): > > > + debug = "input line %d: %s" % (asl[i].lineno, asl[i].line) > > > + > > > + l = asl[i].line > > > + > > > + # skip if not an extract directive > > > + a = len(re.findall(r'ACPI_EXTRACT', l)) > > > + if (not a): > > > + # If not empty, store AML offset. Will be used for sanity checks > > > + # IASL seems to put {}. at random places in the listing. > > > + # Ignore any non-words for the purpose of this test. > > > + m = re.search(r'\w+', l) > > > + if (m): > > > + prev_aml_offset = asl[i].aml_offset > > > + continue > > > + > > > + if (a > 1): > > > + die("Expected at most one ACPI_EXTRACT per line, actual %d" % > a) > > > + > > > + mext = re.search(r''' > > > + ^\s* # leading whitespace > > > + /\*\s* # start C comment > > > + (ACPI_EXTRACT_\w+) # directive: group(1) > > > + \s+ # whitspace separates directive from array > name > > > + (\w+) # array name: group(2) > > > + \s*\*/ # end of C comment > > > + \s*$ # trailing whitespace > > > + ''', l, re.VERBOSE) > > > + if (not mext): > > > + die("Stray ACPI_EXTRACT in input") > > > + > > > + # previous command must have produced some AML, > > > + # otherwise we are in a middle of a block > > > + if (prev_aml_offset == asl[i].aml_offset): > > > + die("ACPI_EXTRACT directive in the middle of a block") > > > + > > > + directive = mext.group(1) > > > + array = mext.group(2) > > > + offset = asl[i].aml_offset > > > + > > > + if (directive == "ACPI_EXTRACT_ALL_CODE"): > > > + if array in output: > > > + die("%s directive used more than once" % directive) > > > + output[array] = aml > > > + continue > > > + if (directive == "ACPI_EXTRACT_NAME_DWORD_CONST"): > > > + offset = aml_name_dword_const(offset) > > > + elif (directive == "ACPI_EXTRACT_NAME_WORD_CONST"): > > > + offset = aml_name_word_const(offset) > > > + elif (directive == "ACPI_EXTRACT_NAME_BYTE_CONST"): > > > + offset = aml_name_byte_const(offset) > > > + elif (directive == "ACPI_EXTRACT_NAME_STRING"): > > > + offset = aml_name_string(offset) > > > + elif (directive == "ACPI_EXTRACT_METHOD_STRING"): > > > + offset = aml_method_string(offset) > > > + elif (directive == "ACPI_EXTRACT_PROCESSOR_START"): > > > + offset = aml_processor_start(offset) > > > + elif (directive == "ACPI_EXTRACT_PROCESSOR_STRING"): > > > + offset = aml_processor_string(offset) > > > + elif (directive == "ACPI_EXTRACT_PROCESSOR_END"): > > > + offset = aml_processor_end(offset) > > > + elif (directive == "ACPI_EXTRACT_PKG_START"): > > > + offset = aml_package_start(offset) > > > + else: > > > + die("Unsupported directive %s" % directive) > > > + > > > + if array not in output: > > > + output[array] = [] > > > + output[array].append(offset) > > > + > > > +debug = "at end of file" > > > + > > > +def get_value_type(maxvalue): > > > + #Use type large enough to fit the table > > > + if (maxvalue >= 0x10000): > > > + return "int" > > > + elif (maxvalue >= 0x100): > > > + return "short" > > > + else: > > > + return "char" > > > + > > > +# Pretty print output > > > +for array in output.keys(): > > > + otype = get_value_type(max(output[array])) > > > + odata = [] > > > + for value in output[array]: > > > + odata.append("0x%x" % value) > > > + sys.stdout.write("unsigned %s %s[] = {\n" % (otype, array)) > > > + sys.stdout.write(",\n".join(odata)) > > > + sys.stdout.write('\n};\n'); > > > diff --git a/tools/firmware/hvmloader/tools/acpi_extract_preprocess.py > b/tools/firmware/hvmloader/tools/acpi_extract_preprocess.py > > > new file mode 100644 > > > index 0000000..4ae364e > > > --- /dev/null > > > +++ b/tools/firmware/hvmloader/tools/acpi_extract_preprocess.py > > > @@ -0,0 +1,41 @@ > > > +#!/usr/bin/python > > > +# Copyright (C) 2011 Red Hat, Inc., Michael S. Tsirkin <mst(a)redhat.com> > > > +# > > > +# This file may be distributed under the terms of the GNU GPLv3 license. > > > + > > > +# Read a preprocessed ASL listing and put each ACPI_EXTRACT > > > +# directive in a comment, to make iasl skip it. > > > +# We also put each directive on a new line, the machinery > > > +# in tools/acpi_extract.py requires this. > > > + > > > +import re; > > > +import sys; > > > +import fileinput; > > > + > > > +def die(diag): > > > + sys.stderr.write("Error: %s\n" % (diag)) > > > + sys.exit(1) > > > + > > > +# Note: () around pattern make split return matched string as part of list > > > +psplit = re.compile(r''' ( > > > + \b # At word boundary > > > + ACPI_EXTRACT_\w+ # directive > > > + \s+ # some whitespace > > > + \w+ # array name > > > + )''', re.VERBOSE); > > > + > > > +lineno = 0 > > > +for line in fileinput.input(): > > > + # line number and debug string to output in case of errors > > > + lineno = lineno + 1 > > > + debug = "input line %d: %s" % (lineno, line.rstrip()) > > > + > > > + s = psplit.split(line); > > > + # The way split works, each odd item is the matching ACPI_EXTRACT > directive. > > > + # Put each in a comment, and on a line by itself. > > > + for i in range(len(s)): > > > + if (i % 2): > > > + sys.stdout.write("\n/* %s */\n" % s[i]) > > > + else: > > > + sys.stdout.write(s[i]) > > > + > > > -- > > > 1.8.3.4 > > > > > > > > > > > > _______________________________________________ > > > Xen-devel mailing list > > > Xen-devel(a)lists.xen.org > > > http://lists.xen.org/xen-devel

6 years, 5 months

[SeaBIOS [PATCH V5 0/2] hw/pci: reserve IO and mem for pci-2-pci bridges with no devices attached

by Marcel Apfelbaum

v4 -> v5 - Addressed Michael S. Tsirkin's comments (patch 2/2): - Open-coded pci_config_is_reserved() method. v3 -> v4: - Addressed Kevin O'Connor's comments: - Refactored a for loop in patch 1/2. - Addressed Michael S. Tsirkin's comments (patch 2/2): - Removed not needed method - Test only base registers (dropped limits tests) - Renamed a helper method - Used 0xFF to test if the memory is reserved - Simplified code in pci_bridge_has_region - I did keep the code that restores base's address as I don't want to modify the registers in a 'query' method. (as replied on the mail thread) v2 -> v3: - Addressed Michael S. Tsirkin's comments: - I/O and Prefetchable Memory are optional. Do not allocate ranges if they are not implemented (2/2). - Note that 2/2 patch can be seen as a separate fix. However, it is related to ranges reservation. v1 -> v2: - Thanks Gerd Hoffmann for the review. - Addressed Michael S. Tsirkin's comments: - Limit capabilities query to 256 iterations, to make sure we don't get into an infinite loop with a broken device. If a pci-2-pci bridge supports hot-plug functionality but there are no devices connected to it, reserve IO/mem in order to be able to attach devices later. Do not waste space, use minimum allowed. Marcel Apfelbaum (2): hw/pci: reserve IO and mem for pci-2-pci bridges with no devices attached hw/pci: check if pci2pci bridges implement optional limit registers src/fw/pciinit.c | 12 +++++------- src/hw/pci.c | 45 +++++++++++++++++++++++++++++++++++++++++++++ src/hw/pci.h | 10 ++++++++++ 3 files changed, 60 insertions(+), 7 deletions(-) -- 1.8.3.1

6 years, 5 months

Re: [SeaBIOS] [Qemu-devel] WfW setup and DOS display.sys hanging

by Dennis Waters

cd roms make bios && make seavgabios && make lgplvgabios ... and it works! Sorry to have troubled you, and thanks for your help :) On 1 May 2014 20:51, Dennis Waters <dennis123321(a)googlemail.com> wrote: > OK this is getting strange or I'm missing something obvious! > > I checked the two affected c source files manually to confirm the changes > had been made. > I've just done a redownload, repatch, and recompile of qemu2.0.0... > > Then I noticed bios.bin & vgabios*.bin are already compiled in the > qemu-2.0.0.tar.bz2 in the "pc-bios" directory. The md5sums of all those > files matches the ones "make install" copies over (even on a patched > source)! > > Am I missing a ./configure flag to force the bios files to recompile? > > Thanks again! > > > On 1 May 2014 20:33, Kevin O'Connor <kevin(a)koconnor.net> wrote: > >> On Thu, May 01, 2014 at 07:56:59PM +0100, Dennis Waters wrote: >> > Hi, >> > >> > Using the latest Qemu2 and Seabios1.7.4 (I've tried stable and git for >> > both) -- I still have this issue with QEMU hanging on display.sys with >> > MSDOS 6.2 >> > I can confirm on all compiled sources the structs already had your >> "PACKED" >> > patch applied >> [...] >> > Scan for VGA option rom >> > Running option rom at c000:0003 >> > Start SeaVGABIOS (version >> > rel-1.7.4-0-g96917a8-20140203_153423-nilsson.home.kraxel.org) >> >> The log indicates the SeaVGABIOS that was used predates the PACKED >> fix. The fix went in after rel-1.7.4 - it's in commit f5ec1e0a. How >> did you confirm the fix was applied? >> >> -Kevin >> > >

6 years, 5 months

Re: [SeaBIOS] [Qemu-devel] WfW setup and DOS display.sys hanging

by Kevin O'Connor

On Thu, May 01, 2014 at 07:56:59PM +0100, Dennis Waters wrote: > Hi, > > Using the latest Qemu2 and Seabios1.7.4 (I've tried stable and git for > both) -- I still have this issue with QEMU hanging on display.sys with > MSDOS 6.2 > I can confirm on all compiled sources the structs already had your "PACKED" > patch applied [...] > Scan for VGA option rom > Running option rom at c000:0003 > Start SeaVGABIOS (version > rel-1.7.4-0-g96917a8-20140203_153423-nilsson.home.kraxel.org) The log indicates the SeaVGABIOS that was used predates the PACKED fix. The fix went in after rel-1.7.4 - it's in commit f5ec1e0a. How did you confirm the fix was applied? -Kevin

6 years, 5 months

Re: [SeaBIOS] [Qemu-devel] WfW setup and DOS display.sys hanging

by Dennis Waters

Hi, Using the latest Qemu2 and Seabios1.7.4 (I've tried stable and git for both) -- I still have this issue with QEMU hanging on display.sys with MSDOS 6.2 I can confirm on all compiled sources the structs already had your "PACKED" patch applied If I swap out vgabios-cirrus.bin for the older lgplvgabios ( savannah.gnu.org/download/vgabios/vgabios-0.7a.cirrus.bin) it boots OK with "-vga cirrus" As requested, below is the debug log from a freshly compiled QEMU2. It sits at 100% cpu when MSDOS tries to load display.sys. I also took the same log file from a successful boot with 0.7a vgabios. Please let me know if you want that posting here too. Thanks for your time! ------ Log file using -chardev stdio,id=seabios -device isa-debugcon,iobase=0x402,chardev=seabios ------ Start bios (version rel-1.7.4-0-g96917a8-20140501_004118-e6540) No Xen hypervisor found. Running on QEMU (i440fx) RamSize: 0x02000000 [cmos] Relocating init from 0x000dfe89 to 0x01fdf190 (size 69036) Found QEMU fw_cfg RamBlock: addr 0x0000000000000000 len 0x0000000002000000 [e820] CPU Mhz=2651 === PCI bus & bridge init === PCI: pci_bios_init_bus_rec bus = 0x0 === PCI device probing === Found 4 PCI devices (max PCI bus is 00) === PCI new allocation pass #1 === PCI: check devices === PCI new allocation pass #2 === PCI: 32: 0000000080000000 - 00000000fec00000 PCI: map device bdf=00:01.1 bar 4, addr 0000c000, size 00000010 [io] PCI: map device bdf=00:02.0 bar 6, addr febe0000, size 00010000 [mem] PCI: map device bdf=00:02.0 bar 1, addr febf0000, size 00001000 [mem] PCI: map device bdf=00:02.0 bar 0, addr fc000000, size 02000000 [prefmem] PCI: init bdf=00:00.0 id=8086:1237 PCI: init bdf=00:01.0 id=8086:7000 PIIX3/PIIX4 init: elcr=00 0c PCI: init bdf=00:01.1 id=8086:7010 PCI: init bdf=00:02.0 id=1013:00b8 PCI: Using 00:02.0 for primary VGA Found 1 cpu(s) max supported 1 cpu(s) MP table addr=0x000f0b70 MPC table addr=0x000f0b80 size=192 SMBIOS ptr=0x000f0b50 table=0x000f0a20 size=304 Scan for VGA option rom Running option rom at c000:0003 Start SeaVGABIOS (version rel-1.7.4-0-g96917a8-20140203_153423-nilsson.home.kraxel.org) enter vga_post: a=00000010 b=0000ffff c=00000000 d=0000ffff ds=0000 es=f000 ss=0000 si=00000000 di=00006580 bp=00000000 sp=00006d1a cs=f000 ip=d256 f=0000 cirrus init cirrus init 2 Attempting to allocate VGA stack via pmm call to f000:d2ce pmm call arg1=0 VGA stack allocated at ed790 Turning on vga text mode console set VGA mode 3 cirrus mode switch regular SeaBIOS (version rel-1.7.4-0-g96917a8-20140501_004118-e6540) Found 1 lpt ports Found 1 serial ports Searching bootorder for: /pci@i0cf8/isa@1/fdc@03f0/floppy@0 ATA controller 1 at 1f0/3f4/0 (irq 14 dev 9) ATA controller 2 at 170/374/0 (irq 15 dev 9) ata0-0: QEMU HARDDISK ATA-7 Hard-Disk (512 MiBytes) Searching bootorder for: /pci@i0cf8/*@1,1/drive@0/disk@0 DVD/CD [ata1-0: QEMU DVD-ROM ATAPI-4 DVD/CD] Searching bootorder for: /pci@i0cf8/*@1,1/drive@1/disk@0 PS2 keyboard initialized All threads complete. Scan for option roms Searching bootorder for: /rom(a)genroms/kvmvapic.bin Searching bootorder for: HALT drive 0x000f09a0: PCHS=1040/16/63 translation=none LCHS=1024/16/63 s=1048576 Running option rom at c900:0003 Space available for UMB: cb800-ec800, f0000-f0940 Returned 65536 bytes of ZoneHigh e820 map has 5 items: 0: 0000000000000000 - 000000000009fc00 = 1 RAM 1: 000000000009fc00 - 00000000000a0000 = 2 RESERVED 2: 00000000000f0000 - 0000000000100000 = 2 RESERVED 3: 0000000000100000 - 0000000002000000 = 1 RAM 4: 00000000fffc0000 - 0000000100000000 = 2 RESERVED enter handle_19: NULL Booting from Hard Disk... Booting from 0000:7c00

6 years, 5 months

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

SeaBIOS May 2014