Interactive Boot Menu for DASD and SCSI Guests on s390x
v8 cover Interactive Boot Menu for DASD and SCSI Guests on s390x
Due to the introduction of the QemuIplParameter block and taking some time to revist some areas, a few chunks of code have been reworked to better align with those changes. I also think the reworkings improve readability. The same functionality is still there, the code just looks a little different (and hopefully looks better).
--- [v8] ---
The tl;dr:
cleaned up some early patches based on review, threw zipl boot option code into its own patch, refactored s390_ipl_set_boot_menu to reflect latest changes.
The "pls explain":
* cleaned up "s390-ccw: refactor boot map table code"
- removed void ptr casting in a couple of spots
* cleaned up "s390-ccw: set cp_receive mask..." patch
- setting the mask consumes a service interrupt, so there is no need for a followup sclp_print
* fixed uitoa based on feedback from v7
* fixed alignment concerns in QemuIplParameters and renamed flags field
- reasoning: necessary; better readability
- s/boot_menu_flags/qipl_flags so this field can be (potentially) used for other purposes in the future
- when setting the boot menu parms in the bios, we take care to mask out the appropriate qipl_flags for boot menu specific flags
- boot menu flags defines are renamed to be prefixed with "QIPL_FLAG_"
- also updated comment above this struct
* cleaned up "s390-ccw: read user input..." patch
- defines for low core external interrupt code addr and clock comparator interrupt code
- take care to make sure buf remains null terminated when passed to read_prompt
* [NEW PATCH] "s390-ccw: use zipl values when no boot menu options are present"
- reasoning: better readability; further explanation of feature
- *nothing new added to this patch series here*
- zipl options flag setting and parsing *moved to* this patch
- this attempts to better explain how these fields are used and how they get parsed
- the commit message gives details on how to set these fields in the zipl configuration file
- the zipl options are only set for CCW type IPL devices (since no other devices actually support it)
* refactored s390_ipl_set_boot_menu
- reasoning: better readability
- the idea is that we should take care to appropriately set the boot menu flags for each IPL device type from the beginning. We should not set certain flags for devices that cannot support certain features (eg SCSI does not support zipl menus, so we should never set the use_zipl_opts flag for SCSI)
- since there are no longer boot menu fields specific to each IPL type, the switch statement is simply used to detect if the IPL device type is capable of a boot menu
- since zipl flags are only set for CCW type IPL devices, I reworked the logic and removed some indentation
* s/menu_check_flags/menu_is_enabled
- reasoning: better readability
- no parameters
- "if menu is enabled" reads better than "if these specific flag bits are set"
* removed menu.h
- reasoning: file not needed; less to maintain
- introduction of qipl and better understanding of zipl options led to this decision. by the end of these changes, this file ended up housing 4 function declarations and no longer seemed necessary
- all menu related function declarations are in s390-ccw.h
- boot menu flags are defined in iplb.h (which aligns with hw/s390x/ipl.h)
--- [Summary] ---
These patches implement a boot menu for ECKD DASD and SCSI guests on s390x. The menu will only appear if the disk has been configured for IPL with the zIPL tool and with the following QEMU command line options:
-boot menu=on[,splash-time=X] and/or -machine loadparm='prompt'
The following must be specified for the device to be IPL'd from:
-device ...,bootindex=1
or via the following libvirt domain xml:
<os>
<bootmenu enable='yes' timeout='X'/>
</os>
or
<disk>
...
<boot order='1' loadparm='PROMPT'/>
</disk>Where X is some positive integer representing time in milliseconds.
<boot order='1' ... > must be specified for the device to be IPL'd from
A loadparm other than 'prompt' will disable the menu and just boot the specified entry.
If no boot options are specified, we will attempt to use the values provided by zipl (ECKD DASD only).
Collin L. Walling (13):
s390-ccw: refactor boot map table code
s390-ccw: refactor eckd_block_num to use CHS
s390-ccw: refactor IPL structs
s390-ccw: update libc
s390-ccw: move auxiliary IPL data to separate location
s390-ccw: parse and set boot menu options
s390-ccw: set up interactive boot menu parameters
s390-ccw: read stage2 boot loader data to find menu
s390-ccw: print zipl boot menu
s390-ccw: read user input for boot index via the SCLP console
s390-ccw: set cp_receive mask only when needed and consume pending
service irqs
s390-ccw: use zipl values when no boot menu options are present
s390-ccw: interactive boot menu for scsi
hw/s390x/ipl.c | 68 ++++++++++++- hw/s390x/ipl.h | 31 +++++- pc-bios/s390-ccw/Makefile | 2 +- pc-bios/s390-ccw/bootmap.c | 184 +++++++++++++++++++++++---------- pc-bios/s390-ccw/bootmap.h | 91 ++++++++++------- pc-bios/s390-ccw/iplb.h | 24 ++++- pc-bios/s390-ccw/libc.c | 88 ++++++++++++++++ pc-bios/s390-ccw/libc.h | 37 ++++++- pc-bios/s390-ccw/main.c | 49 +++++---- pc-bios/s390-ccw/menu.c | 243 ++++++++++++++++++++++++++++++++++++++++++++ pc-bios/s390-ccw/s390-ccw.h | 9 ++ pc-bios/s390-ccw/sclp.c | 39 ++++--- pc-bios/s390-ccw/virtio.c | 2 +- 13 files changed, 741 insertions(+), 126 deletions(-) create mode 100644 pc-bios/s390-ccw/libc.c create mode 100644 pc-bios/s390-ccw/menu.c
Signature · 1 line
2.7.4
PATCH v8 1/13 s390-ccw: refactor boot map table code
Some ECKD bootmap code was using structs designed for SCSI. Even though this works, it confuses readability. Add a new BootMapTable struct to assist with readability in bootmap entry code. Also:
- replace ScsiMbr in ECKD code with appropriate structs
- fix read_block messages to reflect BootMapTable
- fixup ipl_scsi to use BootMapTable (referred to as Program Table)
- defined value for maximum table entries
--- pc-bios/s390-ccw/bootmap.c | 60 +++++++++++++++++++++------------------------- pc-bios/s390-ccw/bootmap.h | 11 ++++++++- 2 files changed, 37 insertions(+), 34 deletions(-)
pc-bios/s390-ccw/bootmap.c +27 -33
@@ -182,24 +182,24 @@ static block_number_t load_eckd_segments(block_number_t blk, uint64_t *address)return block_nr; } -static void run_eckd_boot_script (block_number_t mbr_block_nr) +static void run_eckd_boot_script (block_number_t bmt_block_nr) {int i;unsigned int loadparm =get_loadparm_index ();block_number_t block_nr;uint64_t address; -ScsiMbr *bte = (void *)sec;/* Eckd bootmap table entry */ +BootMapTable *bmt = (void *)sec;BootMapScript *bms = (void *)sec;debug_print_int ("loadparm" , loadparm); -IPL_assert (loadparm <31 ,"loadparm value greater than" +IPL_assert (loadparm <=MAX_TABLE_ENTRIES ,"loadparm value greater than" " maximum number of boot entries allowed" );memset (sec,FREE_SPACE_FILLER ,sizeof (sec)); -read_block (mbr_block_nr, sec,"Cannot read MBR" ); +read_block (bmt_block_nr, sec,"Cannot read Boot Map Table" ); - block_nr =eckd_block_num ((void *)&(bte->blockptr[loadparm])); -IPL_assert (block_nr !=-1 ,"No Boot Map" ); + block_nr =eckd_block_num (&bmt->entry[loadparm]); +IPL_assert (block_nr !=-1 ,"Cannot find Boot Map Table Entry" );memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (block_nr, sec,"Cannot read Boot Map Script" );
@@ -223,7 +223,7 @@ static void ipl_eckd_cdl(void)XEckdMbr *mbr;Ipl2 *ipl2 = (void *)sec;IplVolumeLabel *vlbl = (void *)sec; -block_number_t block_nr; +block_number_t bmt_block_nr;/* we have just read the block #0 and recognized it as "IPL1" */ sclp_print ("CDL\n" );
@@ -238,8 +238,8 @@ static void ipl_eckd_cdl(void)IPL_assert (mbr->dev_type ==DEV_TYPE_ECKD ,"Non-ECKD device type in zIPL section of IPL2 record." ); -/* save pointer to Boot Script */ - block_nr =eckd_block_num ((void *)&(mbr->blockptr)); +/* save pointer to Boot Map Table */ + bmt_block_nr =eckd_block_num (&mbr->blockptr);memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (2 , vlbl,"Cannot read Volume Label at block 2" );
@@ -249,7 +249,7 @@ static void ipl_eckd_cdl(void) "Invalid magic of volser block");print_volser (vlbl->f.volser); -run_eckd_boot_script (block_nr); +run_eckd_boot_script (bmt_block_nr);/* no return */ }
@@ -280,7 +280,7 @@ static void print_eckd_ldl_msg(ECKD_IPL_mode_t mode)static void ipl_eckd_ldl (ECKD_IPL_mode_t mode) { -block_number_t block_nr; +block_number_t bmt_block_nr;BootInfo *bip = (void *)(sec +0x70 );/* BootInfo is MBR for LDL */ if (mode !=ECKD_LDL_UNLABELED ) {
@@ -299,8 +299,10 @@ static void ipl_eckd_ldl(ECKD_IPL_mode_t mode) }verify_boot_info (bip); - block_nr =eckd_block_num ((void *)&(bip->bp.ipl.bm_ptr.eckd.bptr)); -run_eckd_boot_script (block_nr); +/* save pointer to Boot Map Table */ + bmt_block_nr =eckd_block_num ((void *)&bip->bp.ipl.bm_ptr.eckd.bptr); + +run_eckd_boot_script (bmt_block_nr);/* no return */ }
@@ -325,7 +327,7 @@ static void print_eckd_msg(void)static void ipl_eckd (void ) { -ScsiMbr *mbr = (void *)sec; +XEckdMbr *mbr = (void *)sec;LDL_VTOC *vlbl = (void *)sec;print_eckd_msg ();
@@ -449,10 +451,8 @@ static void zipl_run(ScsiBlockPtr *pte)static void ipl_scsi (void ) {ScsiMbr *mbr = (void *)sec; -uint8_t *ns, *ns_end;int program_table_entries =0 ; -const int pte_len =sizeof (ScsiBlockPtr); -ScsiBlockPtr *prog_table_entry =NULL ; +BootMapTable *prog_table = (void *)sec;unsigned int loadparm =get_loadparm_index ();/* Grab the MBR */
@@ -467,34 +467,28 @@ static void ipl_scsi(void)debug_print_int ("MBR Version" , mbr->version_id);IPL_check (mbr->version_id ==1 ,"Unknown MBR layout version, assuming version 1" ); -debug_print_int ("program table" , mbr->blockptr[0 ].blockno); -IPL_assert (mbr->blockptr[0 ].blockno,"No Program Table" ); +debug_print_int ("program table" , mbr->pt.blockno); +IPL_assert (mbr->pt.blockno,"No Program Table" );/* Parse the program table */ -read_block (mbr->blockptr[0 ].blockno, sec, -"Error reading Program Table" ); - +read_block (mbr->pt.blockno, sec,"Error reading Program Table" );IPL_assert (magic_match (sec,ZIPL_MAGIC ),"No zIPL magic in PT" ); -debug_print_int ("loadparm index" , loadparm); - ns_end = sec +virtio_get_block_size (); -for (ns = (sec + pte_len); (ns + pte_len) < ns_end; ns += pte_len) { - prog_table_entry = (ScsiBlockPtr *)ns; -if (!prog_table_entry->blockno) { +while (program_table_entries <=MAX_TABLE_ENTRIES ) { +if (!prog_table->entry[program_table_entries].scsi.blockno) {break ; } - program_table_entries++; -if (program_table_entries == loadparm +1 ) { -break ;/* selected entry found */ - } }debug_print_int ("program table entries" , program_table_entries); -IPL_assert (program_table_entries !=0 ,"Empty Program Table" ); -zipl_run (prog_table_entry);/* no return */ +debug_print_int ("loadparm" , loadparm); +IPL_assert (loadparm <=MAX_TABLE_ENTRIES ,"loadparm value greater than" +" maximum number of boot entries allowed" ); + +zipl_run (&prog_table->entry[loadparm].scsi);/* no return */ }/***********************************************************************
pc-bios/s390-ccw/bootmap.h +10 -1
@@ -53,6 +53,15 @@ typedef union BootMapPointer {ExtEckdBlockPtr xeckd; } __attribute__ ((packed)) BootMapPointer; +#define MAX_TABLE_ENTRIES 30 + +/* aka Program Table */ +typedef struct BootMapTable { +uint8_t magic[4 ]; +uint8_t reserved[12 ]; +BootMapPointer entry[]; +} __attribute__ ((packed))BootMapTable ; + typedef struct ComponentEntry { ScsiBlockPtr data; uint8_t pad[7];
@@ -70,7 +79,7 @@ typedef struct ScsiMbr {uint8_t magic[4 ];uint32_t version_id;uint8_t reserved[8 ]; -ScsiBlockPtr blockptr[]; +ScsiBlockPtr pt;/* block pointer to program table */ } __attribute__ ((packed)) ScsiMbr;#define ZIPL_MAGIC "zIPL"
Signature · 1 line
2.7.4
PATCH v8 2/13 s390-ccw: refactor eckd_block_num to use CHS
In reply to Collin L. WallingAdd new cylinder/head/sector struct. Use it to calculate eckd block numbers instead of a BootMapPointer (which used eckd chs anyway).
--- pc-bios/s390-ccw/bootmap.c | 28 ++++++++++++++-------------- pc-bios/s390-ccw/bootmap.h | 8 ++++++-- 2 files changed, 20 insertions(+), 16 deletions(-)
pc-bios/s390-ccw/bootmap.c +14 -14
@@ -95,32 +95,32 @@ static inline void verify_boot_info(BootInfo *bip) "Bad block size in zIPL section of the 1st record."); } -static block_number_t eckd_block_num (BootMapPointer *p) +static block_number_t eckd_block_num (EckdCHS *chs) {const uint64_t sectors =virtio_get_sectors ();const uint64_t heads =virtio_get_heads (); -const uint64_t cylinder = p->eckd.cylinder - + ((p->eckd.head &0xfff0 ) <<12 ); -const uint64_t head = p->eckd.head &0x000f ; +const uint64_t cylinder = chs->cylinder + + ((chs->head &0xfff0 ) <<12 ); +const uint64_t head = chs->head &0x000f ;const block_number_t block = sectors * heads * cylinder + sectors * head - + p->eckd.sector + + chs->sector -1 ;/* block nr starts with zero */ return block; }static bool eckd_valid_address (BootMapPointer *p) { -const uint64_t head = p->eckd.head &0x000f ; +const uint64_t head = p->eckd.chs.head &0x000f ;if (head >=virtio_get_heads () - || p->eckd.sector >virtio_get_sectors () - || p->eckd.sector <=0 ) { + || p->eckd.chs.sector >virtio_get_sectors () + || p->eckd.chs.sector <=0 ) {return false; }if (!virtio_guessed_disk_nature () && -eckd_block_num (p) >=virtio_get_blocks ()) { +eckd_block_num (&p->eckd.chs) >=virtio_get_blocks ()) {return false; }
@@ -140,7 +140,7 @@ static block_number_t load_eckd_segments(block_number_t blk, uint64_t *address) do { more_data = false; for (j = 0;; j++) { - block_nr = eckd_block_num((void *)&(bprs[j].xeckd)); + block_nr = eckd_block_num(&bprs[j].xeckd.bptr.chs); if (is_null_block_number(block_nr)) { /* end of chunk */ break; }
@@ -198,7 +198,7 @@ static void run_eckd_boot_script(block_number_t bmt_block_nr)memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (bmt_block_nr, sec,"Cannot read Boot Map Table" ); - block_nr =eckd_block_num (&bmt->entry[loadparm]); + block_nr =eckd_block_num (&bmt->entry[loadparm].xeckd.bptr.chs);IPL_assert (block_nr !=-1 ,"Cannot find Boot Map Table Entry" );memset (sec,FREE_SPACE_FILLER ,sizeof (sec));
@@ -206,7 +206,7 @@ static void run_eckd_boot_script(block_number_t bmt_block_nr)for (i =0 ; bms->entry[i].type ==BOOT_SCRIPT_LOAD ; i++) { address = bms->entry[i].address.load_address; - block_nr =eckd_block_num (&(bms->entry[i].blkptr)); + block_nr =eckd_block_num (&bms->entry[i].blkptr.xeckd.bptr.chs); do { block_nr = load_eckd_segments(block_nr, &address);
@@ -239,7 +239,7 @@ static void ipl_eckd_cdl(void) "Non-ECKD device type in zIPL section of IPL2 record.");/* save pointer to Boot Map Table */ - bmt_block_nr =eckd_block_num (&mbr->blockptr); + bmt_block_nr =eckd_block_num (&mbr->blockptr.xeckd.bptr.chs);memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (2 , vlbl,"Cannot read Volume Label at block 2" );
@@ -300,7 +300,7 @@ static void ipl_eckd_ldl(ECKD_IPL_mode_t mode)verify_boot_info (bip);/* save pointer to Boot Map Table */ - bmt_block_nr =eckd_block_num ((void *)&bip->bp.ipl.bm_ptr.eckd.bptr); + bmt_block_nr =eckd_block_num (&bip->bp.ipl.bm_ptr.eckd.bptr.chs);run_eckd_boot_script (bmt_block_nr);/* no return */
pc-bios/s390-ccw/bootmap.h +6 -2
@@ -32,10 +32,14 @@ typedef struct FbaBlockPtr {uint16_t blockct; } __attribute__ ((packed)) FbaBlockPtr; -typedef struct EckdBlockPtr { - uint16_t cylinder; /* cylinder/head/sector is an address of the block */ +typedef struct EckdCHS { +uint16_t cylinder;uint16_t head;uint8_t sector; +} __attribute__ ((packed))EckdCHS ; + +typedef struct EckdBlockPtr { + EckdCHS chs; /* cylinder/head/sector is an address of the block */ uint16_t size; uint8_t count; /* (size_in_blocks-1); * it's 0 for TablePtr, ScriptPtr, and SectionPtr */
Signature · 1 line
2.7.4
PATCH v8 3/13 s390-ccw: refactor IPL structs
In reply to Collin L. WallingECKD DASDs have different IPL structures for CDL and LDL formats. The current Ipl1 and Ipl2 structs follow the CDL format, so we prepend "EckdCdl" to them. Boot info for LDL has been moved to a new struct: EckdLdlIpl1.
--- pc-bios/s390-ccw/bootmap.c | 12 ++++++------ pc-bios/s390-ccw/bootmap.h | 37 +++++++++++++++++++++---------------- 2 files changed, 27 insertions(+), 22 deletions(-)
pc-bios/s390-ccw/bootmap.c +6 -6
@@ -221,7 +221,7 @@ static void run_eckd_boot_script(block_number_t bmt_block_nr)static void ipl_eckd_cdl (void ) {XEckdMbr *mbr; -Ipl2 *ipl2 = (void *)sec; +EckdCdlIpl2 *ipl2 = (void *)sec;IplVolumeLabel *vlbl = (void *)sec;block_number_t bmt_block_nr;
@@ -231,7 +231,7 @@ static void ipl_eckd_cdl(void)memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (1 , ipl2,"Cannot read IPL2 record at block 1" ); - mbr = &ipl2->u.x.mbr; + mbr = &ipl2->mbr;IPL_assert (magic_match (mbr,ZIPL_MAGIC ),"No zIPL section in IPL2 record." );IPL_assert (block_size_ok (mbr->blockptr.xeckd.bptr.size),"Bad block size in zIPL section of IPL2 record." );
@@ -281,7 +281,7 @@ static void print_eckd_ldl_msg(ECKD_IPL_mode_t mode)static void ipl_eckd_ldl (ECKD_IPL_mode_t mode) {block_number_t bmt_block_nr; -BootInfo *bip = (void *)(sec +0x70 );/* BootInfo is MBR for LDL */ +EckdLdlIpl1 *ipl1 = (void *)sec;if (mode !=ECKD_LDL_UNLABELED ) {print_eckd_ldl_msg (mode);
@@ -292,15 +292,15 @@ static void ipl_eckd_ldl(ECKD_IPL_mode_t mode)memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (0 , sec,"Cannot read block 0 to grab boot info." );if (mode ==ECKD_LDL_UNLABELED ) { -if (!magic_match (bip->magic,ZIPL_MAGIC )) { +if (!magic_match (ipl1->bip.magic,ZIPL_MAGIC )) {return ;/* not applicable layout */ }sclp_print ("unlabeled LDL.\n" ); } -verify_boot_info (bip); +verify_boot_info (&ipl1->bip);/* save pointer to Boot Map Table */ - bmt_block_nr =eckd_block_num (&bip->bp.ipl.bm_ptr.eckd.bptr.chs); + bmt_block_nr =eckd_block_num (&ipl1->bip.bp.ipl.bm_ptr.eckd.bptr.chs);run_eckd_boot_script (bmt_block_nr);/* no return */
pc-bios/s390-ccw/bootmap.h +21 -16
@@ -239,22 +239,27 @@ typedef struct BootInfo { /* @ 0x70, record #0 */ } bp; } __attribute__ ((packed)) BootInfo; /* see also XEckdMbr */ -typedef struct Ipl1 { -unsigned char key[4 ];/* == "IPL1" */ -unsigned char data[24 ]; -} __attribute__((packed))Ipl1 ; - -typedef struct Ipl2 { -unsigned char key[4 ];/* == "IPL2" */ -union { -unsigned char data[144 ]; -struct { -unsigned char reserved1[92 -4 ]; -XEckdMbr mbr; -unsigned char reserved2[144 -(92 -4 )-sizeof (XEckdMbr)]; - } x; - } u; -} __attribute__((packed))Ipl2 ; +/* +* Structs for IPL +*/ +#define STAGE2_BLK_CNT_MAX 24/* Stage 1b can load up to 24 blocks */ + +typedef struct EckdCdlIpl1 { +uint8_t key[4 ];/* == "IPL1" */ +uint8_t data[24 ]; +} __attribute__((packed))EckdCdlIpl1 ; + +typedef struct EckdCdlIpl2 { +uint8_t key[4 ];/* == "IPL2" */ +uint8_t reserved0[88 ]; +XEckdMbr mbr; +uint8_t reserved[24 ]; +} __attribute__((packed))EckdCdlIpl2 ; + +typedef struct EckdLdlIpl1 { +uint8_t reserved[112 ]; +BootInfo bip;/* BootInfo is MBR for LDL */ +} __attribute__((packed))EckdLdlIpl1 ; typedef struct IplVolumeLabel { unsigned char key[4]; /* == "VOL1" */
Signature · 1 line
2.7.4
PATCH v8 4/13 s390-ccw: update libc
In reply to Collin L. WallingMoved:
memcmp from bootmap.h to libc.h (renamed from _memcmp) strlen from sclp.c to libc.h (renamed from _strlen)
Added C standard functions:
isdigit
Added non C-standard function:
uitoa
atoui
--- pc-bios/s390-ccw/Makefile | 2 +- pc-bios/s390-ccw/bootmap.c | 4 +-- pc-bios/s390-ccw/bootmap.h | 16 +-------- pc-bios/s390-ccw/libc.c | 88 ++++++++++++++++++++++++++++++++++++++++++++++ pc-bios/s390-ccw/libc.h | 37 +++++++++++++++++-- pc-bios/s390-ccw/main.c | 17 +-------- pc-bios/s390-ccw/sclp.c | 10 +----- 7 files changed, 129 insertions(+), 45 deletions(-) create mode 100644 pc-bios/s390-ccw/libc.c
pc-bios/s390-ccw/Makefile +1 -1
@@ -9,7 +9,7 @@ $(call set-vpath, $(SRC_PATH)/pc-bios/s390-ccw) .PHONY : all clean build-all -OBJECTS = start.o main.o bootmap.o sclp.o virtio.o virtio-scsi.o virtio-blkdev.o +OBJECTS = start.o main.o bootmap.o sclp.o virtio.o virtio-scsi.o virtio-blkdev.o libc.o QEMU_CFLAGS := $(filter -W%, $(QEMU_CFLAGS)) QEMU_CFLAGS += -ffreestanding -fno-delete-null-pointer-checks -msoft-float QEMU_CFLAGS += -march=z900 -fPIE -fno-strict-aliasing
pc-bios/s390-ccw/bootmap.c +2 -2
@@ -506,7 +506,7 @@ static bool is_iso_bc_entry_compatible(IsoBcSection *s) "Failed to read image sector 0");/* Checking bytes 8 - 32 for S390 Linux magic */ -return !_memcmp (magic_sec +8 , linux_s390_magic,24 ); +return !memcmp (magic_sec +8 , linux_s390_magic,24 ); }/* Location of the current sector of the directory */
@@ -635,7 +635,7 @@ static uint32_t find_iso_bc(void)if (vd->type ==VOL_DESC_TYPE_BOOT ) {IsoVdElTorito *et = &vd->vd.boot; -if (!_memcmp (&et->el_torito[0 ], el_torito_magic,32 )) { +if (!memcmp (&et->el_torito[0 ], el_torito_magic,32 )) {return bswap32 (et->bc_offset); } }
pc-bios/s390-ccw/bootmap.h +1 -15
@@ -328,20 +328,6 @@ static inline bool magic_match(const void *data, const void *magic)return *((uint32_t *)data) == *((uint32_t *)magic); } -static inline int _memcmp (const void *s1,const void *s2,size_t n) -{ -int i; -const uint8_t *p1 = s1, *p2 = s2; - -for (i =0 ; i < n; i++) { -if (p1[i] != p2[i]) { -return p1[i] > p2[i] ?1 :-1 ; - } - } - -return 0 ; -} -static inline uint32_t iso_733_to_u32 (uint64_t x) {return (uint32_t )x;
@@ -434,7 +420,7 @@ const uint8_t vol_desc_magic[] = "CD001";static inline bool is_iso_vd_valid (IsoVolDesc *vd) { -return !_memcmp (&vd->ident[0 ], vol_desc_magic,5 ) && +return !memcmp (&vd->ident[0 ], vol_desc_magic,5 ) && vd->version ==0x1 && vd->type <=VOL_DESC_TYPE_PARTITION ; }
pc-bios/s390-ccw/libc.c new file +88 -0
88 lines not on this page · raw message
pc-bios/s390-ccw/libc.h +35 -2
@@ -1,6 +1,8 @@/* * libc-style definitions and functions * +* Copyright (c) 2013 Alexander Graf <[email protected]> +* * This code is free software; you can redistribute it and/or modify it * under the terms of the GNU General Public License as published by the * Free Software Foundation; either version 2 of the License, or (at your
@@ -19,7 +21,7 @@ typedef unsigned long long uint64_t;static inline void *memset (void *s,int c,size_t n) { -int i; +size_t i;unsigned char *p = s;for (i =0 ; i < n; i++) {
@@ -33,7 +35,7 @@ static inline void *memcpy(void *s1, const void *s2, size_t n) {uint8_t *dest = s1;const uint8_t *src = s2; -int i; +size_t i;for (i =0 ; i < n; i++) { dest[i] = src[i];
@@ -42,4 +44,35 @@ static inline void *memcpy(void *s1, const void *s2, size_t n)return s1; } +static inline int memcmp (const void *s1,const void *s2,size_t n) +{ +size_t i; +const uint8_t *p1 = s1, *p2 = s2; + +for (i =0 ; i < n; i++) { +if (p1[i] != p2[i]) { +return p1[i] > p2[i] ?1 :-1 ; + } + } + +return 0 ; +} + +static inline size_t strlen (const char *str) +{ +size_t i; +for (i =0 ; *str; i++) { + str++; + } +return i; +} + +static inline int isdigit (int c) +{ +return (c >='0' ) && (c <='9' ); +} + +uint64_t atoui (const char *str); +char *uitoa (uint64_t num,char *str,size_t len); +#endif
pc-bios/s390-ccw/main.c +1 -16
@@ -40,22 +40,7 @@ void panic(const char *string)unsigned int get_loadparm_index (void ) { -const char *lp = loadparm; -int i; -unsigned int idx =0 ; - -for (i =0 ; i <8 ; i++) { -char c = lp[i]; - -if (c <'0' || c >'9' ) { -break ; - } - - idx *=10 ; - idx += c -'0' ; - } - -return idx; +return atoui (loadparm); }static bool find_dev (Schib *schib,int dev_no)
pc-bios/s390-ccw/sclp.c +1 -9
@@ -65,14 +65,6 @@ void sclp_setup(void)sclp_set_write_mask (); } -static int _strlen (const char *str) -{ -int i; -for (i =0 ; *str; i++) - str++; -return i; -} -long write (int fd,const void *str,size_t len) {WriteEventData *sccb = (void *)_sccb;
@@ -113,7 +105,7 @@ long write(int fd, const void *str, size_t len)void sclp_print (const char *str) { -write (1 , str,_strlen (str)); +write (1 , str,strlen (str)); }void sclp_get_loadparm_ascii (char *loadparm)
Signature · 1 line
2.7.4
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 04/13] s390-ccw: update libc
Quoted 14 lines · Collin L. Walling
Moved:
memcmp frombootmap.htolibc.h(renamed from_memcmp) strlen fromsclp.ctolibc.h(renamed from_strlen)Added C standard functions:
isdigitAdded non C-standard function:
uitoa
atouiSigned-off-by Collin L. Walling <[email protected]>Reviewed-by Janosch Frank <[email protected]>
Quoted 10 lines
--- pc-bios/s390-ccw/Makefile | 2 +- pc-bios/s390-ccw/bootmap.c | 4 +-- pc-bios/s390-ccw/bootmap.h | 16 +-------- pc-bios/s390-ccw/libc.c | 88 ++++++++++++++++++++++++++++++++++++++++++++++ pc-bios/s390-ccw/libc.h | 37 +++++++++++++++++-- pc-bios/s390-ccw/main.c | 17 +-------- pc-bios/s390-ccw/sclp.c | 10 +----- 7 files changed, 129 insertions(+), 45 deletions(-) create mode 100644 pc-bios/s390-ccw/libc.c
PATCH v8 5/13 s390-ccw: move auxiliary IPL data to separate location
In reply to Collin L. WallingThe s390-ccw firmware needs some information in support of the boot process which is not available on the native machine. Examples are the netboot firmware load address and now the boot menu parameters.
While storing that data in unused fields of the IPL parameter block works, that approach could create problems if the parameter block definition should change in the future. Because then a guest could overwrite these fields using the set IPLB diagnose.
In fact the data in question is of more global nature and not really tied to an IPL device, so separating it is rather logical.
This commit introduces a new structure to hold firmware relevant IPL parameters set by QEMU. The data is stored at location 204 (dec) and can contain up to 7 32-bit words. This area is available to programming in the z/Architecture Principles of Operation and can thus safely be used by the firmware until the IPL has completed.
--- hw/s390x/ipl.c | 18 +++++++++++++++++- hw/s390x/ipl.h | 25 +++++++++++++++++++++++-- pc-bios/s390-ccw/iplb.h | 18 ++++++++++++++++-- pc-bios/s390-ccw/main.c | 6 +++++- 4 files changed, 61 insertions(+), 6 deletions(-)
hw/s390x/ipl.c +17 -1
@@ -399,6 +399,21 @@ void s390_reipl_request(void)qemu_system_reset_request (SHUTDOWN_CAUSE_GUEST_RESET ); } +static void s390_ipl_prepare_qipl (S390CPU *cpu) +{ +S390IPLState *ipl =get_ipl_device (); +uint8_t *addr; +uint64_t len =4096 ; + + addr =cpu_physical_memory_map (cpu->env.psa, &len,1 ); +if (!addr || len <QIPL_ADDRESS +sizeof (QemuIplParameters)) { +error_report ("Cannot set QEMU IPL parameters" ); +return ; + } +memcpy (addr +QIPL_ADDRESS , &ipl->qipl,sizeof (QemuIplParameters)); +cpu_physical_memory_unmap (addr, len,1 , len); +} +void s390_ipl_prepare_cpu (S390CPU *cpu) {S390IPLState *ipl =get_ipl_device ();
@@ -418,8 +433,9 @@ void s390_ipl_prepare_cpu(S390CPU *cpu)error_report_err (err);vm_stop (RUN_STATE_INTERNAL_ERROR ); } - ipl->iplb.ccw.netboot_start_addr =cpu_to_be64 (ipl->start_addr); + ipl->qipl.netboot_start_addr =cpu_to_be64 (ipl->start_addr); } +s390_ipl_prepare_qipl (cpu); }static void s390_ipl_reset (DeviceState *dev)
hw/s390x/ipl.h +23 -2
@@ -16,8 +16,7 @@#include "cpu.h" struct IplBlockCcw { -uint64_t netboot_start_addr; -uint8_t reserved0[77 ]; +uint8_t reserved0[85 ];uint8_t ssid;uint16_t devno;uint8_t vm_flags;
@@ -90,6 +89,27 @@ void s390_ipl_prepare_cpu(S390CPU *cpu);IplParameterBlock *s390_ipl_get_iplb (void );void s390_reipl_request (void ); +#define QIPL_ADDRESS 0xcc + +/* +* The QEMU IPL Parameters will be stored at absolute address +* 204 (0xcc) which means it is 32-bit word aligned but not +* double-word aligned. +* Placement of data fields in this area must account for +* their alignment needs. E.g., netboot_start_address must +* have an offset of n * 8 bytes within the struct in order +* to keep it double-word aligned. +* The total size of the struct must never exceed 28 bytes. +* This definition must be kept in sync with the defininition +* in pc-bios/s390-ccw/iplb.h. +*/ +struct QemuIplParameters { +uint8_t reserved1[4 ]; +uint64_t netboot_start_addr; +uint8_t reserved2[16 ]; +}QEMU_PACKED ; +typedef struct QemuIplParameters QemuIplParameters ; +#define TYPE_S390_IPL "s390-ipl"#define S390_IPL (obj) OBJECT_CHECK(S390IPLState, (obj), TYPE_S390_IPL)
@@ -105,6 +125,7 @@ struct S390IPLState {bool iplb_valid;bool reipl_requested;bool netboot; +QemuIplParameters qipl;/*< public >*/ char *kernel;
pc-bios/s390-ccw/iplb.h +16 -2
@@ -13,8 +13,7 @@#define IPLB_H struct IplBlockCcw { -uint64_t netboot_start_addr; -uint8_t reserved0[77 ]; +uint8_t reserved0[85 ];uint8_t ssid;uint16_t devno;uint8_t vm_flags;
@@ -73,6 +72,21 @@ typedef struct IplParameterBlock IplParameterBlock; extern IplParameterBlock iplb __attribute__((__aligned__(PAGE_SIZE))); +#define QIPL_ADDRESS 0xcc + +/* +* This definition must be kept in sync with the defininition +* in hw/s390x/ipl.h +*/ +struct QemuIplParameters { +uint8_t reserved1[4 ]; +uint64_t netboot_start_addr; +uint8_t reserved2[16 ]; +} __attribute__ ((packed)); +typedef struct QemuIplParameters QemuIplParameters ; + +extern QemuIplParameters qipl; +#define S390_IPL_TYPE_FCP 0x00#define S390_IPL_TYPE_CCW 0x02#define S390_IPL_TYPE_QEMU_SCSI 0xff
pc-bios/s390-ccw/main.c +5 -1
@@ -16,6 +16,7 @@ char stack[PAGE_SIZE * 8] __attribute__((__aligned__(PAGE_SIZE)));static SubChannelId blk_schid = { .one =1 }; IplParameterBlock iplb __attribute__((__aligned__(PAGE_SIZE)));static char loadparm[8 ] = {0 ,0 ,0 ,0 ,0 ,0 ,0 ,0 }; +QemuIplParameters qipl;/* * Priniciples of Operations (SA22-7832-09) chapter 17 requires that
@@ -81,6 +82,7 @@ static void virtio_setup(void)uint16_t dev_no;char ldp[] ="LOADPARM=[________]\n" ;VDev *vdev =virtio_get_device (); +QemuIplParameters *early_qipl = (QemuIplParameters *)QIPL_ADDRESS ;/* * We unconditionally enable mss support. In every sane configuration,
@@ -93,6 +95,8 @@ static void virtio_setup(void)memcpy (ldp +10 , loadparm,8 );sclp_print (ldp); +memcpy (&qipl, early_qipl,sizeof (QemuIplParameters)); +if (store_iplb (&iplb)) {switch (iplb.pbt) {case S390_IPL_TYPE_CCW :
@@ -127,7 +131,7 @@ static void virtio_setup(void)if (virtio_get_device_type () ==VIRTIO_ID_NET ) {sclp_print ("Network boot device detected\n" ); - vdev->netboot_start_addr = iplb.ccw.netboot_start_addr; + vdev->netboot_start_addr = qipl.netboot_start_addr; }else {virtio_blk_setup_device (blk_schid);
Signature · 1 line
2.7.4
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 05/13] s390-ccw: move auxiliary IPL data to separate location
Quoted 83 lines · Collin L. Walling
The s390-ccw firmware needs some information in support of the boot process which is not available on the native machine. Examples are the netboot firmware load address and now the boot menu parameters.
While storing that data in unused fields of the IPL parameter block works, that approach could create problems if the parameter block definition should change in the future. Because then a guest could overwrite these fields using the set IPLB diagnose.
In fact the data in question is of more global nature and not really tied to an IPL device, so separating it is rather logical.
This commit introduces a new structure to hold firmware relevant IPL parameters set by QEMU. The data is stored at location 204 (dec) and can contain up to 7 32-bit words. This area is available to programming in the z/Architecture Principles of Operation and can thus safely be used by the firmware until the IPL has completed.
Signed-off-by Viktor Mihajlovski <[email protected]>Signed-off-by Collin L. Walling <[email protected]>--- hw/s390x/ipl.c | 18 +++++++++++++++++- hw/s390x/ipl.h | 25 +++++++++++++++++++++++-- pc-bios/s390-ccw/iplb.h | 18 ++++++++++++++++-- pc-bios/s390-ccw/main.c | 6 +++++- 4 files changed, 61 insertions(+), 6 deletions(-)hw/s390x/ipl.c +17
-1@@ -399,6 +399,21 @@ void s390_reipl_request(void)qemu_system_reset_request (SHUTDOWN_CAUSE_GUEST_RESET ); } +static void s390_ipl_prepare_qipl (S390CPU *cpu) +{ +S390IPLState *ipl =get_ipl_device (); +uint8_t *addr; +uint64_t len =4096 ; + + addr =cpu_physical_memory_map (cpu->env.psa, &len,1 ); +if (!addr || len <QIPL_ADDRESS +sizeof (QemuIplParameters)) { +error_report ("Cannot set QEMU IPL parameters" ); +return ; + } +memcpy (addr +QIPL_ADDRESS , &ipl->qipl,sizeof (QemuIplParameters)); +cpu_physical_memory_unmap (addr, len,1 , len); +} +void s390_ipl_prepare_cpu (S390CPU *cpu) {S390IPLState *ipl =get_ipl_device ();@@ -418,8 +433,9 @@ void s390_ipl_prepare_cpu(S390CPU *cpu)error_report_err (err);vm_stop (RUN_STATE_INTERNAL_ERROR ); } - ipl->iplb.ccw.netboot_start_addr =cpu_to_be64 (ipl->start_addr); + ipl->qipl.netboot_start_addr =cpu_to_be64 (ipl->start_addr); } +s390_ipl_prepare_qipl (cpu); }static void s390_ipl_reset (DeviceState *dev)hw/s390x/ipl.h +1
-2@@ -16,8 +16,7 @@#include "cpu.h" struct IplBlockCcw { -uint64_t netboot_start_addr; -uint8_t reserved0[77 ]; +uint8_t reserved0[85 ];uint8_t ssid;uint16_t devno;uint8_t vm_flags;@@ -90,6 +89,27 @@ void s390_ipl_prepare_cpu(S390CPU *cpu);IplParameterBlock *s390_ipl_get_iplb (void );void s390_reipl_request (void );
hw/s390x/ipl.h void s390_ipl_prepare_cpu(S390CPU *cpu); · line 101
Quoted 10 lines
hw/s390x/ipl.h +10
-0+#define QIPL_ADDRESS 0xcc + +/* +* The QEMU IPL Parameters will be stored at absolute address +* 204 (0xcc) which means it is 32-bit word aligned but not +* double-word aligned. +* Placement of data fields in this area must account for +* their alignment needs. E.g., netboot_start_address must +* have an offset of n * 8 bytes within the struct in order +* to keep it double-word aligned.
Should that rather be "4 + n * 8" instead of "n * 8" ?
Apart from that, patch looks good to me now, so once you've fixed the comment (if necessary):
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 05/13] s390-ccw: move auxiliary IPL data to separate location
Quoted 84 lines · Thomas Huth
Quoted 83 lines · Collin L. Walling
The s390-ccw firmware needs some information in support of the boot process which is not available on the native machine. Examples are the netboot firmware load address and now the boot menu parameters.
While storing that data in unused fields of the IPL parameter block works, that approach could create problems if the parameter block definition should change in the future. Because then a guest could overwrite these fields using the set IPLB diagnose.
In fact the data in question is of more global nature and not really tied to an IPL device, so separating it is rather logical.
This commit introduces a new structure to hold firmware relevant IPL parameters set by QEMU. The data is stored at location 204 (dec) and can contain up to 7 32-bit words. This area is available to programming in the z/Architecture Principles of Operation and can thus safely be used by the firmware until the IPL has completed.
Signed-off-by Viktor Mihajlovski <[email protected]>Signed-off-by Collin L. Walling <[email protected]>--- hw/s390x/ipl.c | 18 +++++++++++++++++- hw/s390x/ipl.h | 25 +++++++++++++++++++++++-- pc-bios/s390-ccw/iplb.h | 18 ++++++++++++++++-- pc-bios/s390-ccw/main.c | 6 +++++- 4 files changed, 61 insertions(+), 6 deletions(-)hw/s390x/ipl.c +17
-1@@ -399,6 +399,21 @@ void s390_reipl_request(void)qemu_system_reset_request (SHUTDOWN_CAUSE_GUEST_RESET ); } +static void s390_ipl_prepare_qipl (S390CPU *cpu) +{ +S390IPLState *ipl =get_ipl_device (); +uint8_t *addr; +uint64_t len =4096 ; + + addr =cpu_physical_memory_map (cpu->env.psa, &len,1 ); +if (!addr || len <QIPL_ADDRESS +sizeof (QemuIplParameters)) { +error_report ("Cannot set QEMU IPL parameters" ); +return ; + } +memcpy (addr +QIPL_ADDRESS , &ipl->qipl,sizeof (QemuIplParameters)); +cpu_physical_memory_unmap (addr, len,1 , len); +} +void s390_ipl_prepare_cpu (S390CPU *cpu) {S390IPLState *ipl =get_ipl_device ();@@ -418,8 +433,9 @@ void s390_ipl_prepare_cpu(S390CPU *cpu)error_report_err (err);vm_stop (RUN_STATE_INTERNAL_ERROR ); } - ipl->iplb.ccw.netboot_start_addr =cpu_to_be64 (ipl->start_addr); + ipl->qipl.netboot_start_addr =cpu_to_be64 (ipl->start_addr); } +s390_ipl_prepare_qipl (cpu); }static void s390_ipl_reset (DeviceState *dev)hw/s390x/ipl.h +1
-2@@ -16,8 +16,7 @@#include "cpu.h" struct IplBlockCcw { -uint64_t netboot_start_addr; -uint8_t reserved0[77 ]; +uint8_t reserved0[85 ];uint8_t ssid;uint16_t devno;uint8_t vm_flags;@@ -90,6 +89,27 @@ void s390_ipl_prepare_cpu(S390CPU *cpu);IplParameterBlock *s390_ipl_get_iplb (void );void s390_reipl_request (void );
Quoted 12 lines
hw/s390x/ipl.h void s390_ipl_prepare_cpu(S390CPU *cpu); · line 101
Quoted 10 lines
hw/s390x/ipl.h +10
-0+#define QIPL_ADDRESS 0xcc + +/* +* The QEMU IPL Parameters will be stored at absolute address +* 204 (0xcc) which means it is 32-bit word aligned but not +* double-word aligned. +* Placement of data fields in this area must account for +* their alignment needs. E.g., netboot_start_address must +* have an offset of n * 8 bytes within the struct in order +* to keep it double-word aligned. Should that rather be "4 + n * 8" instead of "n * 8" ?
I wonder if I ever get that comment right. You're correct of course.
Quoted 6 lines
Apart from that, patch looks good to me now, so once you've fixed the comment (if necessary):
Reviewed-by Thomas Huth <[email protected]>
Signature · 2 lines
Regards, Viktor Mihajlovski
PATCH v8 6/13 s390-ccw: parse and set boot menu options
In reply to Collin L. WallingSet boot menu options for an s390 guest and store them in the iplb. These options are set via the QEMU command line option:
-boot menu=on|off[,splash-time=X]
or via the libvirt domain xml:
<os>
<bootmenu enable='yes|no' timeout='X'/>
</os>Where X represents some positive integer representing milliseconds.
Any value set for loadparm will override all boot menu options. If loadparm=PROMPT, then the menu will be enabled without a timeout.
--- hw/s390x/ipl.c | 44 ++++++++++++++++++++++++++++++++++++++++++++ hw/s390x/ipl.h | 9 +++++++-- pc-bios/s390-ccw/iplb.h | 9 +++++++-- 3 files changed, 58 insertions(+), 4 deletions(-)
hw/s390x/ipl.c +44 -0
@@ -23,6 +23,9 @@#include "hw/s390x/ebcdic.h" #include "ipl.h" #include "qemu/error-report.h" +#include "qemu/config-file.h" +#include "qemu/cutils.h" +#include "qemu/option.h" #define KERN_IMAGE_START 0x010000UL#define KERN_PARM_AREA 0x010480UL
@@ -219,6 +222,46 @@ static Property s390_ipl_properties[] = { DEFINE_PROP_END_OF_LIST(), }; +static void s390_ipl_set_boot_menu (S390IPLState *ipl) +{ +QemuOptsList *plist =qemu_find_opts ("boot-opts" ); +QemuOpts *opts =QTAILQ_FIRST (&plist->head); +uint8_t *flags = &ipl->qipl.qipl_flags; +uint32_t *timeout = &ipl->qipl.boot_menu_timeout; +const char *tmp; +unsigned long splash_time =0 ; + +switch (ipl->iplb.pbt) { +case S390_IPL_TYPE_CCW : +break ; +default : +error_report ("boot menu is not supported for this device type." ); +return ; + } + +if (!boot_menu) { +return ; + } + + *flags |=QIPL_FLAG_BM_OPTS_CMD ; + + tmp =qemu_opt_get (opts,"splash-time" ); + +if (tmp &&qemu_strtoul (tmp,NULL ,10 , &splash_time)) { +error_report ("splash-time is invalid, forcing it to 0." ); + *timeout =0 ; +return ; + } + +if (splash_time >0xffffffff ) { +error_report ("splash-time is too large, forcing it to max value." ); + *timeout =0xffffffff ; +return ; + } + + *timeout =cpu_to_be32 (splash_time); +} +static bool s390_gen_initial_iplb (S390IPLState *ipl) {DeviceState *dev_st;
@@ -435,6 +478,7 @@ void s390_ipl_prepare_cpu(S390CPU *cpu) } ipl->qipl.netboot_start_addr =cpu_to_be64 (ipl->start_addr); } +s390_ipl_set_boot_menu (ipl);s390_ipl_prepare_qipl (cpu); }
hw/s390x/ipl.h +7 -2
@@ -91,6 +91,9 @@ void s390_reipl_request(void);#define QIPL_ADDRESS 0xcc +/* Boot Menu flags */ +#define QIPL_FLAG_BM_OPTS_CMD 0x80 +/* * The QEMU IPL Parameters will be stored at absolute address * 204 (0xcc) which means it is 32-bit word aligned but not
@@ -104,9 +107,11 @@ void s390_reipl_request(void);* in pc-bios/s390-ccw/iplb.h. */ struct QemuIplParameters { -uint8_t reserved1[4 ]; +uint8_t qipl_flags; +uint8_t reserved1[3 ];uint64_t netboot_start_addr; -uint8_t reserved2[16 ]; +uint32_t boot_menu_timeout; +uint8_t reserved2[12 ]; }QEMU_PACKED ;typedef struct QemuIplParameters QemuIplParameters ;
pc-bios/s390-ccw/iplb.h +7 -2
@@ -74,14 +74,19 @@ extern IplParameterBlock iplb __attribute__((__aligned__(PAGE_SIZE)));#define QIPL_ADDRESS 0xcc +/* Boot Menu flags */ +#define QIPL_FLAG_BM_OPTS_CMD 0x80 +/* * This definition must be kept in sync with the defininition * in hw/s390x/ipl.h */ struct QemuIplParameters { -uint8_t reserved1[4 ]; +uint8_t qipl_flags; +uint8_t reserved1[3 ];uint64_t netboot_start_addr; -uint8_t reserved2[16 ]; +uint32_t boot_menu_timeout; +uint8_t reserved2[12 ]; } __attribute__ ((packed));typedef struct QemuIplParameters QemuIplParameters ;
Signature · 1 line
2.7.4
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 06/13] s390-ccw: parse and set boot menu options
Quoted 19 lines · Collin L. Walling
Set boot menu options for an s390 guest and store them in the iplb. These options are set via the QEMU command line option:
-boot menu=on|off[,splash-time=X]
or via the libvirt domain xml:
<os> <bootmenu enable='yes|no' timeout='X'/> </os>Where X represents some positive integer representing milliseconds.
Any value set for loadparm will override all boot menu options. If loadparm=PROMPT, then the menu will be enabled without a timeout.
Quoted 7 lines
Signed-off-by Collin L. Walling <[email protected]>Reviewed-by Janosch Frank <[email protected]>--- hw/s390x/ipl.c | 44 ++++++++++++++++++++++++++++++++++++++++++++ hw/s390x/ipl.h | 9 +++++++-- pc-bios/s390-ccw/iplb.h | 9 +++++++-- 3 files changed, 58 insertions(+), 4 deletions(-)
PATCH v8 7/13 s390-ccw: set up interactive boot menu parameters
In reply to Collin L. WallingReads boot menu flag and timeout values from the iplb and sets the respective fields for the menu.
--- pc-bios/s390-ccw/Makefile | 2 +- pc-bios/s390-ccw/main.c | 24 ++++++++++++++++++++++++ pc-bios/s390-ccw/menu.c | 27 +++++++++++++++++++++++++++ pc-bios/s390-ccw/s390-ccw.h | 4 ++++ 4 files changed, 56 insertions(+), 1 deletion(-) create mode 100644 pc-bios/s390-ccw/menu.c
pc-bios/s390-ccw/Makefile +1 -1
@@ -9,7 +9,7 @@ $(call set-vpath, $(SRC_PATH)/pc-bios/s390-ccw) .PHONY : all clean build-all -OBJECTS = start.o main.o bootmap.o sclp.o virtio.o virtio-scsi.o virtio-blkdev.o libc.o +OBJECTS = start.o main.o bootmap.o sclp.o virtio.o virtio-scsi.o virtio-blkdev.o libc.o menu.o QEMU_CFLAGS := $(filter -W%, $(QEMU_CFLAGS)) QEMU_CFLAGS += -ffreestanding -fno-delete-null-pointer-checks -msoft-float QEMU_CFLAGS += -march=z900 -fPIE -fno-strict-aliasing
pc-bios/s390-ccw/main.c +24 -0
@@ -18,6 +18,9 @@ IplParameterBlock iplb __attribute__((__aligned__(PAGE_SIZE)));static char loadparm[8 ] = {0 ,0 ,0 ,0 ,0 ,0 ,0 ,0 };QemuIplParameters qipl; +#define LOADPARM_PROMPT "PROMPT " +#define LOADPARM_EMPTY "........" +/* * Priniciples of Operations (SA22-7832-09) chapter 17 requires that * a subsystem-identification is at 184-187 and bytes 188-191 are zero
@@ -74,6 +77,26 @@ static bool find_dev(Schib *schib, int dev_no)return false; } +static void menu_setup (void ) +{ +if (memcmp (loadparm,LOADPARM_PROMPT ,8 ) ==0 ) { +menu_set_parms (QIPL_FLAG_BM_OPTS_CMD ,0 ); +return ; + } + +/* If loadparm was set to any other value, then do not enable menu */ +if (memcmp (loadparm,LOADPARM_EMPTY ,8 ) !=0 ) { +return ; + } + +switch (iplb.pbt) { +case S390_IPL_TYPE_CCW : +menu_set_parms (qipl.qipl_flags &QIPL_FLAG_BM_OPTS_CMD , + qipl.boot_menu_timeout); +return ; + } +} +static void virtio_setup (void ) {Schib schib;
@@ -117,6 +140,7 @@ static void virtio_setup(void)default :panic ("List-directed IPL not supported yet!\n" ); } +menu_setup (); }else {for (ssid =0 ; ssid <0x3 ; ssid++) { blk_schid.ssid = ssid;
pc-bios/s390-ccw/menu.c new file +27 -0
@@ -0,0 +1,27 @@ +/* +* QEMU S390 Interactive Boot Menu +* +* Copyright 2018 IBM Corp. +* Author: Collin L. Walling <[email protected]> +* +* This work is licensed under the terms of the GNU GPL, version 2 or (at +* your option) any later version. See the COPYING file in the top-level +* directory. +*/ + +#include "libc.h" +#include "s390-ccw.h" + +static uint8_t flag; +static uint64_t timeout; + +void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout) +{ + flag = boot_menu_flag; + timeout = boot_menu_timeout; +} + +bool menu_is_enabled (void ) +{ +return flag; +}
pc-bios/s390-ccw/s390-ccw.h +4 -0
@@ -84,6 +84,10 @@ ulong get_second(void);/* bootmap.c */ void zipl_load (void ); +/* menu.c */ +void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout); +bool menu_is_enabled (void ); +static inline void fill_hex (char *out,unsigned char val) {const char hex[] ="0123456789abcdef" ;
Signature · 1 line
2.7.4
PATCH v8 8/13 s390-ccw: read stage2 boot loader data to find menu
In reply to Collin L. WallingRead the stage2 boot loader data block-by-block. We scan the current block for the string "zIPL" to detect the start of the boot menu banner. We then load the adjacent blocks (previous block and next block) to account for the possibility of menu data spanning multiple blocks.
--- pc-bios/s390-ccw/bootmap.c | 94 ++++++++++++++++++++++++++++++++++++++++++--- pc-bios/s390-ccw/bootmap.h | 23 ++++++++++- pc-bios/s390-ccw/menu.c | 5 +++ pc-bios/s390-ccw/s390-ccw.h | 1 + 4 files changed, 116 insertions(+), 7 deletions(-)
pc-bios/s390-ccw/bootmap.c +89 -5
@@ -83,6 +83,10 @@ static void jump_to_IPL_code(uint64_t address)static unsigned char _bprs[8 *1024 ];/* guessed "max" ECKD sector size */ static const int max_bprs_entries =sizeof (_bprs) /sizeof (ExtEckdBlockPtr); +static uint8_t _s2[MAX_SECTOR_SIZE * 3] __attribute__((__aligned__(PAGE_SIZE))); +static void *s2_prev_blk = _s2; +static void *s2_cur_blk = _s2 +MAX_SECTOR_SIZE ; +static void *s2_next_blk = _s2 +MAX_SECTOR_SIZE *2 ;static inline void verify_boot_info (BootInfo *bip) {
@@ -182,7 +186,77 @@ static block_number_t load_eckd_segments(block_number_t blk, uint64_t *address)return block_nr; } -static void run_eckd_boot_script (block_number_t bmt_block_nr) +static bool find_zipl_boot_menu_banner (int *offset) +{ +int i; + +/* Menu banner starts with "zIPL" */ +for (i =0 ; i <virtio_get_block_size () -4 ; i++) { +if (magic_match (s2_cur_blk + i,ZIPL_MAGIC_EBCDIC )) { + *offset = i; +return true; + } + } + +return false; +} + +static int eckd_get_boot_menu_index (block_number_t s1b_block_nr) +{ +block_number_t cur_block_nr; +block_number_t prev_block_nr =0 ; +block_number_t next_block_nr =0 ; +EckdStage1b *s1b = (void *)sec; +int banner_offset; +int i; + +/* Get Stage1b data */ +memset (sec,FREE_SPACE_FILLER ,sizeof (sec)); +read_block (s1b_block_nr, s1b,"Cannot read stage1b boot loader" ); + +memset (_s2,FREE_SPACE_FILLER ,sizeof (_s2)); + +/* Get Stage2 data */ +for (i =0 ; i <STAGE2_BLK_CNT_MAX ; i++) { + cur_block_nr =eckd_block_num (&s1b->seek[i].chs); + +if (!cur_block_nr) { +break ; + } + +read_block (cur_block_nr, s2_cur_blk,"Cannot read stage2 boot loader" ); + +if (find_zipl_boot_menu_banner (&banner_offset)) { +/* +* Load the adjacent blocks to account for the +* possibility of menu data spanning multiple blocks. +*/ +if (prev_block_nr) { +read_block (prev_block_nr, s2_prev_blk, +"Cannot read stage2 boot loader" ); + } + +if (i +1 <STAGE2_BLK_CNT_MAX ) { + next_block_nr =eckd_block_num (&s1b->seek[i +1 ].chs); + } + +if (next_block_nr) { +read_block (next_block_nr, s2_next_blk, +"Cannot read stage2 boot loader" ); + } + +return menu_get_zipl_boot_index (s2_cur_blk + banner_offset); + } + + prev_block_nr = cur_block_nr; + } + +sclp_print ("No zipl boot menu data found. Booting default entry." ); +return 0 ; +} + +static void run_eckd_boot_script (block_number_t bmt_block_nr, +block_number_t s1b_block_nr) {int i;unsigned int loadparm =get_loadparm_index ();
@@ -191,6 +265,10 @@ static void run_eckd_boot_script(block_number_t bmt_block_nr)BootMapTable *bmt = (void *)sec;BootMapScript *bms = (void *)sec; +if (menu_is_enabled ()) { + loadparm =eckd_get_boot_menu_index (s1b_block_nr); + } +debug_print_int ("loadparm" , loadparm);IPL_assert (loadparm <=MAX_TABLE_ENTRIES ,"loadparm value greater than" " maximum number of boot entries allowed" );
@@ -223,7 +301,7 @@ static void ipl_eckd_cdl(void)XEckdMbr *mbr;EckdCdlIpl2 *ipl2 = (void *)sec;IplVolumeLabel *vlbl = (void *)sec; -block_number_t bmt_block_nr; +block_number_t bmt_block_nr, s1b_block_nr;/* we have just read the block #0 and recognized it as "IPL1" */ sclp_print ("CDL\n" );
@@ -241,6 +319,9 @@ static void ipl_eckd_cdl(void)/* save pointer to Boot Map Table */ bmt_block_nr =eckd_block_num (&mbr->blockptr.xeckd.bptr.chs); +/* save pointer to Stage1b Data */ + s1b_block_nr =eckd_block_num (&ipl2->stage1.seek[0 ].chs); +memset (sec,FREE_SPACE_FILLER ,sizeof (sec));read_block (2 , vlbl,"Cannot read Volume Label at block 2" ); IPL_assert(magic_match(vlbl->key, VOL1_MAGIC),
@@ -249,7 +330,7 @@ static void ipl_eckd_cdl(void) "Invalid magic of volser block");print_volser (vlbl->f.volser); -run_eckd_boot_script (bmt_block_nr); +run_eckd_boot_script (bmt_block_nr, s1b_block_nr);/* no return */ }
@@ -280,7 +361,7 @@ static void print_eckd_ldl_msg(ECKD_IPL_mode_t mode)static void ipl_eckd_ldl (ECKD_IPL_mode_t mode) { -block_number_t bmt_block_nr; +block_number_t bmt_block_nr, s1b_block_nr;EckdLdlIpl1 *ipl1 = (void *)sec;if (mode !=ECKD_LDL_UNLABELED ) {
@@ -302,7 +383,10 @@ static void ipl_eckd_ldl(ECKD_IPL_mode_t mode)/* save pointer to Boot Map Table */ bmt_block_nr =eckd_block_num (&ipl1->bip.bp.ipl.bm_ptr.eckd.bptr.chs); -run_eckd_boot_script (bmt_block_nr); +/* save pointer to Stage1b Data */ + s1b_block_nr =eckd_block_num (&ipl1->stage1.seek[0 ].chs); + +run_eckd_boot_script (bmt_block_nr, s1b_block_nr);/* no return */ }
pc-bios/s390-ccw/bootmap.h +21 -2
@@ -87,6 +87,7 @@ typedef struct ScsiMbr { } __attribute__ ((packed)) ScsiMbr;#define ZIPL_MAGIC "zIPL" +#define ZIPL_MAGIC_EBCDIC "\xa9\xc9\xd7\xd3"#define IPL1_MAGIC "\xc9\xd7\xd3\xf1"/* == "IPL1" in EBCDIC */ #define IPL2_MAGIC "\xc9\xd7\xd3\xf2"/* == "IPL2" in EBCDIC */ #define VOL1_MAGIC "\xe5\xd6\xd3\xf1"/* == "VOL1" in EBCDIC */
@@ -249,15 +250,33 @@ typedef struct EckdCdlIpl1 {uint8_t data[24 ]; } __attribute__((packed)) EckdCdlIpl1; +typedef struct EckdSeekArg { +uint16_t pad; +EckdCHS chs; +uint8_t pad2; +} __attribute__ ((packed))EckdSeekArg ; + +typedef struct EckdStage1b { +uint8_t reserved[32 *STAGE2_BLK_CNT_MAX ]; +struct EckdSeekArg seek[STAGE2_BLK_CNT_MAX ]; +uint8_t unused[64 ]; +} __attribute__ ((packed))EckdStage1b ; + +typedef struct EckdStage1 { +uint8_t reserved[72 ]; +struct EckdSeekArg seek[2 ]; +} __attribute__ ((packed))EckdStage1 ; +typedef struct EckdCdlIpl2 {uint8_t key[4 ];/* == "IPL2" */ -uint8_t reserved0[88 ]; +struct EckdStage1 stage1;XEckdMbr mbr;uint8_t reserved[24 ]; } __attribute__((packed))EckdCdlIpl2 ;typedef struct EckdLdlIpl1 { -uint8_t reserved[112 ]; +uint8_t reserved[24 ]; +struct EckdStage1 stage1;BootInfo bip;/* BootInfo is MBR for LDL */ } __attribute__((packed))EckdLdlIpl1 ;
pc-bios/s390-ccw/menu.c +5 -0
@@ -15,6 +15,11 @@static uint8_t flag;static uint64_t timeout; +int menu_get_zipl_boot_index (const char *menu_data) +{ +return 0 ;/* implemented next patch */ +} +void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout) { flag = boot_menu_flag;
pc-bios/s390-ccw/s390-ccw.h +1 -0
@@ -87,6 +87,7 @@ void zipl_load(void);/* menu.c */ void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout);bool menu_is_enabled (void ); +int menu_get_zipl_boot_index (const char *menu_data);static inline void fill_hex (char *out,unsigned char val) {
Signature · 1 line
2.7.4
PATCH v8 9/13 s390-ccw: print zipl boot menu
In reply to Collin L. WallingWhen the boot menu options are present and the guest's disk has been configured by the zipl tool, then the user will be presented with an interactive boot menu with labeled entries. An example of what the menu might look like:
zIPL v1.37.1-build-20170714 interactive boot menu.
0. default (linux-4.13.0)
1. linux-4.13.0
2. performance
3. kvm
--- pc-bios/s390-ccw/menu.c | 33 ++++++++++++++++++++++++++++++++- 1 file changed, 32 insertions(+), 1 deletion(-)
pc-bios/s390-ccw/menu.c +32 -1
@@ -15,11 +15,42 @@static uint8_t flag;static uint64_t timeout; -int menu_get_zipl_boot_index (const char *menu_data) +static int get_boot_index (int entries) {return 0 ;/* implemented next patch */ } +static void zipl_println (const char *data,size_t len) +{ +char buf[len +2 ]; + +ebcdic_to_ascii (data, buf, len); + buf[len] ='\n' ; + buf[len +1 ] ='\0' ; + +sclp_print (buf); +} + +int menu_get_zipl_boot_index (const char *menu_data) +{ +size_t len; +int entries; + +/* Print and count all menu items, including the banner */ +for (entries =0 ; *menu_data; entries++) { + len =strlen (menu_data); +zipl_println (menu_data, len); + menu_data += len +1 ; + +if (entries <2 ) { +sclp_print ("\n" ); + } + } + +sclp_print ("\n" ); +return get_boot_index (entries -1 );/* subtract 1 to exclude banner */ +} +void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout) { flag = boot_menu_flag;
Signature · 1 line
2.7.4
PATCH v8 10/13 s390-ccw: read user input for boot index via the SCLP console
In reply to Collin L. WallingImplements an sclp_read function to capture input from the console and a wrapper function that handles parsing certain characters and adding input to a buffer. The input is checked for any erroneous values and is handled appropriately.
A prompt will persist until input is entered or the timeout expires (if one was set). Example:
Please choose (default will boot in 10 seconds):
Correct input will boot the respective boot index. If the user's input is empty, 0, or if the timeout expires, then the default zipl entry will be chosen. If the input is within the range of available boot entries, then the selection will be booted. Any erroneous input will cancel the timeout and re-prompt the user.
--- pc-bios/s390-ccw/menu.c | 149 +++++++++++++++++++++++++++++++++++++++++++- pc-bios/s390-ccw/s390-ccw.h | 2 + pc-bios/s390-ccw/sclp.c | 19 ++++++ pc-bios/s390-ccw/virtio.c | 2 +- 4 files changed, 170 insertions(+), 2 deletions(-)
pc-bios/s390-ccw/menu.c +148 -1
@@ -12,12 +12,159 @@#include "libc.h" #include "s390-ccw.h" +#define KEYCODE_NO_INP '\0' +#define KEYCODE_ESCAPE '\033' +#define KEYCODE_BACKSP '\177' +#define KEYCODE_ENTER '\r' + +#define TOD_CLOCK_MILLISECOND 0x3e8000 + +#define LOW_CORE_EXTERNAL_INT_ADDR 0x86 +#define CLOCK_COMPARATOR_INT 0X1004 +static uint8_t flag;static uint64_t timeout; +static inline void enable_clock_int (void ) +{ +uint64_t tmp =0 ; + + asmvolatile ( +"stctg 0,0,%0\n" +"oi 6+%0, 0x8\n" +"lctlg 0,0,%0" + : :"Q" (tmp) :"memory" + ); +} + +static inline void disable_clock_int (void ) +{ +uint64_t tmp =0 ; + + asmvolatile ( +"stctg 0,0,%0\n" +"ni 6+%0, 0xf7\n" +"lctlg 0,0,%0" + : :"Q" (tmp) :"memory" + ); +} + +static inline void set_clock_comparator (uint64_t time) +{ + asmvolatile ("sckc %0" : :"Q" (time)); +} + +static inline bool check_clock_int (void ) +{ +uint16_t *code = (uint16_t *)LOW_CORE_EXTERNAL_INT_ADDR ; + +consume_sclp_int (); + +return *code ==CLOCK_COMPARATOR_INT ; +} + +static int read_prompt (char *buf,size_t len) +{ +char inp[2 ] = {}; +uint8_t idx =0 ; +uint64_t time; + +if (timeout) { + time =get_clock () + timeout *TOD_CLOCK_MILLISECOND ; +set_clock_comparator (time); +enable_clock_int (); + timeout =0 ; + } + +while (!check_clock_int ()) { + +sclp_read (inp,1 );/* Process only one character at a time */ + +switch (inp[0 ]) { +case KEYCODE_NO_INP : +case KEYCODE_ESCAPE : +continue ; +case KEYCODE_BACKSP : +if (idx >0 ) { + buf[--idx] =0 ; +sclp_print ("\b \b" ); + } +continue ; +case KEYCODE_ENTER : +disable_clock_int (); +return idx; +default : +/* Echo input and add to buffer */ +if (idx < len) { + buf[idx++] = inp[0 ]; +sclp_print (inp); + } + } + } + +disable_clock_int (); + *buf =0 ; + +return 0 ; +} + +static int get_index (void ) +{ +char buf[11 ]; +int len; +int i; + +memset (buf,0 ,sizeof (buf)); + + len =read_prompt (buf,sizeof (buf) -1 ); + +/* If no input, boot default */ +if (len ==0 ) { +return 0 ; + } + +/* Check for erroneous input */ +for (i =0 ; i < len; i++) { +if (!isdigit (buf[i])) { +return -1 ; + } + } + +return atoui (buf); +} + +static void boot_menu_prompt (bool retry) +{ +char tmp[11 ]; + +if (retry) { +sclp_print ("\nError: undefined configuration" +"\nPlease choose:\n" ); + }else if (timeout >0 ) { +sclp_print ("Please choose (default will boot in " ); +sclp_print (uitoa (timeout /1000 , tmp,sizeof (tmp))); +sclp_print (" seconds):\n" ); + }else { +sclp_print ("Please choose:\n" ); + } +} +static int get_boot_index (int entries) { -return 0 ;/* implemented next patch */ +int boot_index; +bool retry = false; +char tmp[5 ]; + +do { +boot_menu_prompt (retry); + boot_index =get_index (); + retry = true; + }while (boot_index <0 || boot_index >= entries); + +sclp_print ("\nBooting entry #" ); +sclp_print (uitoa (boot_index, tmp,sizeof (tmp))); + +return boot_index; }static void zipl_println (const char *data,size_t len)
pc-bios/s390-ccw/s390-ccw.h +2 -0
@@ -71,6 +71,7 @@ unsigned int get_loadparm_index(void);void sclp_print (const char *string);void sclp_setup (void );void sclp_get_loadparm_ascii (char *loadparm); +int sclp_read (char *str,size_t count);/* virtio.c */ unsigned long virtio_load_direct(ulong rec_list1, ulong rec_list2,
@@ -79,6 +80,7 @@ bool virtio_is_supported(SubChannelId schid);void virtio_blk_setup_device (SubChannelId schid);int virtio_read (ulong sector,void *load_addr);int enable_mss_facility (void ); +u64 get_clock (void );ulong get_second (void );/* bootmap.c */
pc-bios/s390-ccw/sclp.c +19 -0
@@ -119,3 +119,22 @@ void sclp_get_loadparm_ascii(char *loadparm)ebcdic_to_ascii ((char *) sccb->loadparm, loadparm,8 ); } } + +int sclp_read (char *str,size_t count) +{ +ReadEventData *sccb = (void *)_sccb; +char *buf = (char *)(&sccb->ebh) +7 ; + +/* If count exceeds max buffer size, then restrict it to the max size */ +if (count >SCCB_SIZE -8 ) { + count =SCCB_SIZE -8 ; + } + + sccb->h.length =SCCB_SIZE ; + sccb->h.function_code =SCLP_UNCONDITIONAL_READ ; + +sclp_service_call (SCLP_CMD_READ_EVENT_DATA , sccb); +memcpy (str, buf, count); + +return sccb->ebh.length -7 ; +}
pc-bios/s390-ccw/virtio.c +1 -1
@@ -176,7 +176,7 @@ void vring_send_buf(VRing *vr, void *p, int len, int flags) } } -static u64 get_clock (void ) +u64 get_clock (void ) {u64 r;
Signature · 1 line
2.7.4
PATCH v8 11/13 s390-ccw: set cp_receive mask only when needed and consume pending service irqs
In reply to Collin L. WallingIt is possible while waiting for multiple types of external interrupts that we might have pending irqs remaining between irq consumption and irq-type disabling. Those interrupts could potentially propagate to the guest after IPL completes and cause unwanted behavior.
As it is today, the SCLP will only recognize write events that are enabled by the control program's send and receive masks. To limit the window for, and prevent further irqs from, ASCII console events (specifically keystrokes), we should only enable the control program's receive mask when we need it.
While we're at it, remove assignment of the (non control program) send and receive masks, as those are actually set by the SCLP.
--- pc-bios/s390-ccw/menu.c | 5 +++++ pc-bios/s390-ccw/s390-ccw.h | 1 + pc-bios/s390-ccw/sclp.c | 10 ++++------ 3 files changed, 10 insertions(+), 6 deletions(-)
pc-bios/s390-ccw/menu.c +5 -0
@@ -11,6 +11,7 @@#include "libc.h" #include "s390-ccw.h" +#include "sclp.h" #define KEYCODE_NO_INP '\0'#define KEYCODE_ESCAPE '\033'
@@ -116,8 +117,12 @@ static int get_index(void)memset (buf,0 ,sizeof (buf)); +sclp_set_write_mask (SCLP_EVENT_MASK_MSG_ASCII ,SCLP_EVENT_MASK_MSG_ASCII ); + len =read_prompt (buf,sizeof (buf) -1 ); +sclp_set_write_mask (0 ,SCLP_EVENT_MASK_MSG_ASCII ); +/* If no input, boot default */ if (len ==0 ) {return 0 ;
pc-bios/s390-ccw/s390-ccw.h +1 -0
@@ -69,6 +69,7 @@ unsigned int get_loadparm_index(void);/* sclp.c */ void sclp_print (const char *string); +void sclp_set_write_mask (uint32_t receive_mask,uint32_t send_mask);void sclp_setup (void );void sclp_get_loadparm_ascii (char *loadparm);int sclp_read (char *str,size_t count);
pc-bios/s390-ccw/sclp.c +4 -6
@@ -46,23 +46,21 @@ static int sclp_service_call(unsigned int command, void *sccb)return 0 ; } -static void sclp_set_write_mask (void ) +void sclp_set_write_mask (uint32_t receive_mask,uint32_t send_mask) {WriteEventMask *sccb = (void *)_sccb; sccb->h.length =sizeof (WriteEventMask); sccb->mask_length =sizeof (unsigned int ); - sccb->receive_mask =SCLP_EVENT_MASK_MSG_ASCII ; - sccb->cp_receive_mask =SCLP_EVENT_MASK_MSG_ASCII ; - sccb->send_mask =SCLP_EVENT_MASK_MSG_ASCII ; - sccb->cp_send_mask =SCLP_EVENT_MASK_MSG_ASCII ; + sccb->cp_receive_mask = receive_mask; + sccb->cp_send_mask = send_mask;sclp_service_call (SCLP_CMD_WRITE_EVENT_MASK , sccb); }void sclp_setup (void ) { -sclp_set_write_mask (); +sclp_set_write_mask (0 ,SCLP_EVENT_MASK_MSG_ASCII ); }long write (int fd,const void *str,size_t len)
Signature · 1 line
2.7.4
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 11/13] s390-ccw: set cp_receive mask only when needed and consume pending service irqs
Quoted 15 lines · Collin L. Walling
It is possible while waiting for multiple types of external interrupts that we might have pending irqs remaining between irq consumption and irq-type disabling. Those interrupts could potentially propagate to the guest after IPL completes and cause unwanted behavior.
As it is today, the SCLP will only recognize write events that are enabled by the control program's send and receive masks. To limit the window for, and prevent further irqs from, ASCII console events (specifically keystrokes), we should only enable the control program's receive mask when we need it.
While we're at it, remove assignment of the (non control program) send and receive masks, as those are actually set by the SCLP.
Quoted 6 lines
Signed-off-by Collin L. Walling <[email protected]>--- pc-bios/s390-ccw/menu.c | 5 +++++ pc-bios/s390-ccw/s390-ccw.h | 1 + pc-bios/s390-ccw/sclp.c | 10 ++++------ 3 files changed, 10 insertions(+), 6 deletions(-)
PATCH v8 12/13 s390-ccw: use zipl values when no boot menu options are present
In reply to Collin L. WallingIf no boot menu options are present, then flag the boot menu to use the zipl options that were set in the zipl configuration file (and stored on disk by zipl). These options are found at some offset prior to the start of the zipl boot menu banner. The zipl timeout value is limited to a 16-bit unsigned integer and stored as seconds, so we take care to convert it to milliseconds in order to conform to the rest of the boot menu functionality. This is limited to CCW devices.
For reference, the zipl configuration file uses the following fields in the menu section:
prompt=1 enable the boot menu
timeout=X set the timeout to X seconds
To explicitly disregard any boot menu options, then menu=off or <bootmenu enable='no' ... /> must be specified.
--- hw/s390x/ipl.c | 5 +++++ hw/s390x/ipl.h | 1 + pc-bios/s390-ccw/iplb.h | 1 + pc-bios/s390-ccw/main.c | 3 ++- pc-bios/s390-ccw/menu.c | 14 ++++++++++++++ 5 files changed, 23 insertions(+), 1 deletion(-)
hw/s390x/ipl.c +5 -0
@@ -233,6 +233,11 @@ static void s390_ipl_set_boot_menu(S390IPLState *ipl)switch (ipl->iplb.pbt) {case S390_IPL_TYPE_CCW : +/* In the absence of -boot menu, use zipl parameters */ +if (!qemu_opt_get (opts,"menu" )) { + *flags |=QIPL_FLAG_BM_OPTS_ZIPL ; +return ; + }break ;default :error_report ("boot menu is not supported for this device type." );
hw/s390x/ipl.h +1 -0
@@ -93,6 +93,7 @@ void s390_reipl_request(void);/* Boot Menu flags */ #define QIPL_FLAG_BM_OPTS_CMD 0x80 +#define QIPL_FLAG_BM_OPTS_ZIPL 0x40/* * The QEMU IPL Parameters will be stored at absolute address
pc-bios/s390-ccw/iplb.h +1 -0
@@ -76,6 +76,7 @@ extern IplParameterBlock iplb __attribute__((__aligned__(PAGE_SIZE)));/* Boot Menu flags */ #define QIPL_FLAG_BM_OPTS_CMD 0x80 +#define QIPL_FLAG_BM_OPTS_ZIPL 0x40/* * This definition must be kept in sync with the defininition
pc-bios/s390-ccw/main.c +2 -1
@@ -20,6 +20,7 @@ QemuIplParameters qipl;#define LOADPARM_PROMPT "PROMPT "#define LOADPARM_EMPTY "........" +#define BOOT_MENU_FLAG_MASK (QIPL_FLAG_BM_OPTS_CMD | QIPL_FLAG_BM_OPTS_ZIPL)/* * Priniciples of Operations (SA22-7832-09) chapter 17 requires that
@@ -91,7 +92,7 @@ static void menu_setup(void)switch (iplb.pbt) {case S390_IPL_TYPE_CCW : -menu_set_parms (qipl.qipl_flags &QIPL_FLAG_BM_OPTS_CMD , +menu_set_parms (qipl.qipl_flags &BOOT_MENU_FLAG_MASK , qipl.boot_menu_timeout);return ; }
pc-bios/s390-ccw/menu.c +14 -0
@@ -18,6 +18,10 @@#define KEYCODE_BACKSP '\177'#define KEYCODE_ENTER '\r' +/* Offsets from zipl fields to zipl banner start */ +#define ZIPL_TIMEOUT_OFFSET 138 +#define ZIPL_FLAG_OFFSET 140 +#define TOD_CLOCK_MILLISECOND 0x3e8000#define LOW_CORE_EXTERNAL_INT_ADDR 0x86
@@ -187,6 +191,16 @@ int menu_get_zipl_boot_index(const char *menu_data) {size_t len;int entries; +uint16_t zipl_flag = *(uint16_t *)(menu_data -ZIPL_FLAG_OFFSET ); +uint16_t zipl_timeout = *(uint16_t *)(menu_data -ZIPL_TIMEOUT_OFFSET ); + +if (flag ==QIPL_FLAG_BM_OPTS_ZIPL ) { +if (!zipl_flag) { +return 0 ;/* Boot default */ + } +/* zipl stores timeout as seconds */ + timeout = zipl_timeout *1000 ; + }/* Print and count all menu items, including the banner */ for (entries =0 ; *menu_data; entries++) {
Signature · 1 line
2.7.4
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 12/13] s390-ccw: use zipl values when no boot menu options are present
Quoted 18 lines · Collin L. Walling
If no boot menu options are present, then flag the boot menu to use the zipl options that were set in the zipl configuration file (and stored on disk by zipl). These options are found at some offset prior to the start of the zipl boot menu banner. The zipl timeout value is limited to a 16-bit unsigned integer and stored as seconds, so we take care to convert it to milliseconds in order to conform to the rest of the boot menu functionality. This is limited to CCW devices.
For reference, the zipl configuration file uses the following fields in the menu section:
prompt=1 enable the boot menu
timeout=X set the timeout to X secondsTo explicitly disregard any boot menu options, then menu=off or <bootmenu enable='no' ... /> must be specified.
Quoted 8 lines
Signed-off-by Collin L. Walling <[email protected]>--- hw/s390x/ipl.c | 5 +++++ hw/s390x/ipl.h | 1 + pc-bios/s390-ccw/iplb.h | 1 + pc-bios/s390-ccw/main.c | 3 ++- pc-bios/s390-ccw/menu.c | 14 ++++++++++++++ 5 files changed, 23 insertions(+), 1 deletion(-)
PATCH v8 13/13 s390-ccw: interactive boot menu for scsi
In reply to Collin L. WallingInteractive boot menu for scsi. This follows a similar procedure as the interactive menu for eckd dasd. An example follows:
s390x Enumerated Boot Menu.
3 entries detected. Select from index 0 to 2.
--- hw/s390x/ipl.c | 1 + pc-bios/s390-ccw/bootmap.c | 4 ++++ pc-bios/s390-ccw/main.c | 1 + pc-bios/s390-ccw/menu.c | 14 ++++++++++++++ pc-bios/s390-ccw/s390-ccw.h | 1 + 5 files changed, 21 insertions(+)
hw/s390x/ipl.c +1 -0
@@ -238,6 +238,7 @@ static void s390_ipl_set_boot_menu(S390IPLState *ipl) *flags |=QIPL_FLAG_BM_OPTS_ZIPL ;return ; } +case S390_IPL_TYPE_QEMU_SCSI :break ;default :error_report ("boot menu is not supported for this device type." );
pc-bios/s390-ccw/bootmap.c +4 -0
@@ -568,6 +568,10 @@ static void ipl_scsi(void)debug_print_int ("program table entries" , program_table_entries);IPL_assert (program_table_entries !=0 ,"Empty Program Table" ); +if (menu_is_enabled ()) { + loadparm =menu_get_enum_boot_index (program_table_entries); + } +debug_print_int ("loadparm" , loadparm);IPL_assert (loadparm <=MAX_TABLE_ENTRIES ,"loadparm value greater than" " maximum number of boot entries allowed" );
pc-bios/s390-ccw/main.c +1 -0
@@ -92,6 +92,7 @@ static void menu_setup(void)switch (iplb.pbt) {case S390_IPL_TYPE_CCW : +case S390_IPL_TYPE_QEMU_SCSI :menu_set_parms (qipl.qipl_flags &BOOT_MENU_FLAG_MASK , qipl.boot_menu_timeout);return ;
pc-bios/s390-ccw/menu.c +14 -0
@@ -217,6 +217,20 @@ int menu_get_zipl_boot_index(const char *menu_data)return get_boot_index (entries -1 );/* subtract 1 to exclude banner */ } +int menu_get_enum_boot_index (int entries) +{ +char tmp[4 ]; + +sclp_print ("s390x Enumerated Boot Menu.\n\n" ); + +sclp_print (uitoa (entries, tmp,sizeof (tmp))); +sclp_print (" entries detected. Select from boot index 0 to " ); +sclp_print (uitoa (entries -1 , tmp,sizeof (tmp))); +sclp_print (".\n\n" ); + +return get_boot_index (entries); +} +void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout) { flag = boot_menu_flag;
pc-bios/s390-ccw/s390-ccw.h +1 -0
@@ -91,6 +91,7 @@ void zipl_load(void);void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout);bool menu_is_enabled (void );int menu_get_zipl_boot_index (const char *menu_data); +int menu_get_enum_boot_index (int entries);static inline void fill_hex (char *out,unsigned char val) {
Signature · 1 line
2.7.4
Re: [Qemu-devel] [qemu-s390x] [PATCH v8 13/13] s390-ccw: interactive boot menu for scsi
Quoted 22 lines · Collin L. Walling
Interactive boot menu for scsi. This follows a similar procedure as the interactive menu for eckd dasd. An example follows:
s390x Enumerated Boot Menu.
3 entries detected. Select from index 0 to 2.
Signed-off-by Collin L. Walling <[email protected]>Reviewed-by Thomas Huth <[email protected]>--- hw/s390x/ipl.c | 1 + pc-bios/s390-ccw/bootmap.c | 4 ++++ pc-bios/s390-ccw/main.c | 1 + pc-bios/s390-ccw/menu.c | 14 ++++++++++++++ pc-bios/s390-ccw/s390-ccw.h | 1 + 5 files changed, 21 insertions(+)hw/s390x/ipl.c +0
-0
hw/s390x/ipl.c static void s390_ipl_set_boot_menu(S390IPLState *ipl) · line 242
Quoted 5 lines
hw/s390x/ipl.c +1
-0@@ -238,6 +238,7 @@ static void s390_ipl_set_boot_menu(S390IPLState *ipl) *flags |=QIPL_FLAG_BM_OPTS_ZIPL ;return ; } +case S390_IPL_TYPE_QEMU_SCSI :break ;
It's not a real bug, but I'm pretty sure this will cause Coverity to report an issue of unintended switch-case fall through. Could you please add a break before the new case label?
(Since all other patches look fine to me, this could maybe also be fixed when the patches are picked up, so no need to respin the whole patch series just because of this)
Thomas
Series
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
If you start QEMU manually without a bootindex, the -boot menu=on is ignored if no drive has a bootindex.
For example:
-drive file=/dev/dasda,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on does work
-drive file=/dev/dasda -boot menu=on
does not
instead it prints:
qemu-system-s390x: boot menu is not supported for this device type.
and the boots up the default entry.
Quoted 170 lines · Collin L. Walling
Due to the introduction of the QemuIplParameter block and taking some time to revist some areas, a few chunks of code have been reworked to better align with those changes. I also think the reworkings improve readability. The same functionality is still there, the code just looks a little different (and hopefully looks better).
--- [v8] ---
The tl;dr:
cleaned up some early patches based on review, threw zipl boot option code into its own patch, refactored
s390_ipl_set_boot_menuto reflect latest changes.The "pls explain":
* cleaned up "s390-ccw: refactor boot map table code"
- removed void ptr casting in a couple of spots
* cleaned up "s390-ccw: set
cp_receivemask..." patch- setting the mask consumes a service interrupt, so there is no need for a followup
sclp_print* fixed uitoa based on feedback from v7
* fixed alignment concerns in QemuIplParameters and renamed flags field
- reasoning: necessary; better readability
- s/boot_menu_flags/qipl_flags so this field can be (potentially) used for other purposes in the future
- when setting the boot menu parms in the bios, we take care to mask out the appropriate
qipl_flagsfor boot menu specific flags- boot menu flags defines are renamed to be prefixed with "
QIPL_FLAG_"- also updated comment above this struct
* cleaned up "s390-ccw: read user input..." patch
- defines for low core external interrupt code addr and clock comparator interrupt code
- take care to make sure buf remains null terminated when passed to
read_prompt* [NEW PATCH] "s390-ccw: use zipl values when no boot menu options are present"
- reasoning: better readability; further explanation of feature
- *nothing new added to this patch series here*
- zipl options flag setting and parsing *moved to* this patch
- this attempts to better explain how these fields are used and how they get parsed
- the commit message gives details on how to set these fields in the zipl configuration file
- the zipl options are only set for CCW type IPL devices (since no other devices actually support it)
* refactored
s390_ipl_set_boot_menu- reasoning: better readability
- the idea is that we should take care to appropriately set the boot menu flags for each IPL device type from the beginning. We should not set certain flags for devices that cannot support certain features (eg SCSI does not support zipl menus, so we should never set the
use_zipl_optsflag for SCSI)- since there are no longer boot menu fields specific to each IPL type, the switch statement is simply used to detect if the IPL device type is capable of a boot menu
- since zipl flags are only set for CCW type IPL devices, I reworked the logic and removed some indentation
* s/menu_check_flags/menu_is_enabled
- reasoning: better readability
- no parameters
- "if menu is enabled" reads better than "if these specific flag bits are set"
* removed
menu.h- reasoning: file not needed; less to maintain
- introduction of qipl and better understanding of zipl options led to this decision. by the end of these changes, this file ended up housing 4 function declarations and no longer seemed necessary
- all menu related function declarations are in
s390-ccw.h- boot menu flags are defined in
iplb.h(which aligns withhw/s390x/ipl.h)--- [Summary] ---
These patches implement a boot menu for ECKD DASD and SCSI guests on s390x. The menu will only appear if the disk has been configured for IPL with the zIPL tool and with the following QEMU command line options:
-boot menu=on[,splash-time=X] and/or -machine loadparm='prompt'
The following must be specified for the device to be IPL'd from:
-device ...,bootindex=1or via the following libvirt domain xml:
<os> <bootmenu enable='yes' timeout='X'/> </os> or <disk> ... <boot order='1' loadparm='PROMPT'/> </disk>Where X is some positive integer representing time in milliseconds.
<boot order='1' ... > must be specified for the device to be IPL'd from
A loadparm other than 'prompt' will disable the menu and just boot the specified entry.
If no boot options are specified, we will attempt to use the values provided by zipl (ECKD DASD only).
Collin L. Walling (13):
s390-ccw: refactor boot map table code
s390-ccw: refactoreckd_block_numto use CHS
s390-ccw: refactor IPL structs
s390-ccw: update libc
s390-ccw: move auxiliary IPL data to separate location
s390-ccw: parse and set boot menu options
s390-ccw: set up interactive boot menu parameters
s390-ccw: read stage2 boot loader data to find menu
s390-ccw: print zipl boot menu
s390-ccw: read user input for boot index via the SCLP console
s390-ccw: setcp_receivemask only when needed and consume pending
service irqs
s390-ccw: use zipl values when no boot menu options are present
s390-ccw: interactive boot menu for scsihw/s390x/ipl.c | 68 ++++++++++++- hw/s390x/ipl.h | 31 +++++- pc-bios/s390-ccw/Makefile | 2 +- pc-bios/s390-ccw/bootmap.c | 184 +++++++++++++++++++++++---------- pc-bios/s390-ccw/bootmap.h | 91 ++++++++++------- pc-bios/s390-ccw/iplb.h | 24 ++++- pc-bios/s390-ccw/libc.c | 88 ++++++++++++++++ pc-bios/s390-ccw/libc.h | 37 ++++++- pc-bios/s390-ccw/main.c | 49 +++++---- pc-bios/s390-ccw/menu.c | 243 ++++++++++++++++++++++++++++++++++++++++++++ pc-bios/s390-ccw/s390-ccw.h | 9 ++ pc-bios/s390-ccw/sclp.c | 39 ++++--- pc-bios/s390-ccw/virtio.c | 2 +- 13 files changed, 741 insertions(+), 126 deletions(-) create mode 100644 pc-bios/s390-ccw/libc.c create mode 100644 pc-bios/s390-ccw/menu.c
Quoted 17 lines · Christian Borntraeger
Series
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
If you start QEMU manually without a bootindex, the -boot menu=on is ignored if no drive has a bootindex.
For example:
-drive file=/dev/dasda,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on does work
Quoted 8 lines
-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently in the BIOS up to QEMU (find the first defined virtio disk and select it as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB. While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]
Signature · 2 lines
Regards, Viktor Mihajlovski
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 21 lines
Quoted 21 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
If you start QEMU manually without a bootindex, the -boot menu=on is ignored if no drive has a bootindex.
For example:
-drive file=/dev/dasda,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on does work
-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.
Quoted 12 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently in the BIOS up to QEMU (find the first defined virtio disk and select it as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB. While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]
I can look into it. Theoretically, the easier fix should just involve parsing all of the -device commands and looking for a "bootindex=1" field. The Qemu options code already handles a bulk of this work, so it's just a matter of putting it all together.
Shall I whip something up and post what I have as a reply to this email chain?
Signature · 1 line
- Collin L Walling
Quoted 40 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 21 lines
Quoted 21 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
If you start QEMU manually without a bootindex, the -boot menu=on is ignored if no drive has a bootindex.
For example:
-drive file=/dev/dasda,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on does work
-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 12 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently in the BIOS up to QEMU (find the first defined virtio disk and select it as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB. While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]
Quoted 6 lines
I can look into it. Theoretically, the easier fix should just involve parsing all of the -device commands and looking for a "bootindex=1" field. The Qemu options code already handles a bulk of this work, so it's just a matter of putting it all together.
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool s390_gen_initial_iplb(S390IPLState *ipl)
{
DeviceState *dev_st;
dev_st = get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default disk.
Quoted 48 lines · Christian Borntraeger
Quoted 38 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 20 lines
Quoted 20 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
If you start QEMU manually without a bootindex, the -boot menu=on is ignored if no drive has a bootindex.
For example:
-drive file=/dev/dasda,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on does work
-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 12 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently in the BIOS up to QEMU (find the first defined virtio disk and select it as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB. While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]Quoted 6 lines
I can look into it. Theoretically, the easier fix should just involve parsing all of the -device commands and looking for a "bootindex=1" field. The Qemu options code already handles a bulk of this work, so it's just a matter of putting it all together.
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
Quoted 10 lines
static bool
s390_gen_initial_iplb(S390IPLState *ipl){ DeviceState *dev_st; dev_st = get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default disk.
Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu options present means we fallback to using zipl values for CCW being tied into the switch statement is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once I have something good.
Signature · 1 line
- Collin L Walling
Quoted 61 lines · Collin L. Walling
Quoted 53 lines · Christian Borntraeger
Quoted 39 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 21 lines
Quoted 21 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
If you start QEMU manually without a bootindex, the -boot menu=on is ignored if no drive has a bootindex.
For example:
-drive file=/dev/dasda,if=none,id=d1 -device
virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on
does work-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 12 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently in the BIOS up to QEMU (find the first defined virtio disk and select it as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB. While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]Quoted 6 lines
I can look into it. Theoretically, the easier fix should just involve parsing all of the -device commands and looking for a "bootindex=1" field. The Qemu options code already handles a bulk of this work, so it's just a matter of putting it all together.
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool
s390_gen_initial_iplb(S390IPLState *ipl){
DeviceState*dev_st;Quoted 7 lines
dev_st=get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default
disk.
Quoted 5 lines
Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu options present means we fallback to using zipl values for CCW being tied into the switch statement is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once I have something good.
This should do the trick (this can also be squished painlessly into 6/13 if desired):
From dea9f22429cb3e4ccc81980974adc9f55ead87bf Mon Sep 17 00:00:00 2001
From: "Collin L. Walling" <[email protected]>
Date: Thu, 22 Feb 2018 14:26:14 -0500
Subject: [PATCH] s390-ccw: set boot menu options iff a bootindex was provided
The boot menu options should be set iff a device was specified with bootindex=1, otherwise an error message will appear claiming that the boot menu is not available for a particular device type. This can be misleading.
To amend this, only set the boot menu options when
-device ...,bootindex=1
is provded on the command line for a valid IPL device. The absence of a bootindex provided with -boot menu=on will display an error message, and -boot menu=off (or absence of -boot options) will return silently.
---
hw/s390x/ipl.c | 8 ++++++++
1 file changed, 8 insertions(+)
hw/s390x/ipl.c +0 -0
@@ -231,6 +231,14 @@ static void s390_ipl_set_boot_menu(S390IPLState *ipl)
const char *tmp;
unsigned long splash_time = 0;
+ if (!get_boot_device(0)) {
+ if (boot_menu) {
+ error_report("boot menu requires a bootindex to be specified for " + "the IPL device.");
+ }
+ return;
+ }
+
switch (ipl->iplb.pbt) {
case S390_IPL_TYPE_CCW:
/* In the absence of -boot menu, use zipl parameters */
Signature · 4 lines
2.7.4 -- - Collin L Walling
Quoted 79 lines · Collin L. Walling
Quoted 70 lines · Collin L. Walling
Quoted 62 lines · Christian Borntraeger
Quoted 52 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 23 lines
Quoted 23 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer
scenario (no libvirt)).If you start QEMU manually without a bootindex, the -boot menu=on is ignored
if no drive has a bootindex.For example:
-drive file=/dev/dasda,if=none,id=d1 -device
virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on
does work-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 15 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently
in the BIOS up to QEMU (find the first defined virtio disk and select it
as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB.
While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]I can look into it. Theoretically, the easier fix should just involve parsing all
of the -device commands and looking for a "bootindex=1" field. The Qemu options
code already handles a bulk of this work, so it's just a matter of putting it all
together.Quoted 2 lines
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool
s390_gen_initial_iplb(S390IPLState *ipl){
DeviceState*dev_st;Quoted 7 lines
dev_st=get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default
disk.Quoted 8 lines
Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu
options present means we fallback to using zipl values for CCW being tied into the switch statement
is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once
I have something good.
Quoted 2 lines
This should do the trick (this can also be squished painlessly into 6/13 if desired)
Patch looks fine to me. I can either take it directly like this, or in case you have to respin (depends on the problem that Christian reported with the Ubuntu guest), I'm also fine if you squash it into an earlier patch instead.
Thomas
Quoted 80 lines · Thomas Huth
Quoted 79 lines · Collin L. Walling
Quoted 78 lines · Collin L. Walling
Quoted 62 lines · Christian Borntraeger
Quoted 52 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 23 lines
Quoted 23 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer
scenario (no libvirt)).If you start QEMU manually without a bootindex, the -boot menu=on is ignored
if no drive has a bootindex.For example:
-drive file=/dev/dasda,if=none,id=d1 -device
virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on
does work-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 15 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently
in the BIOS up to QEMU (find the first defined virtio disk and select it
as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB.
While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]I can look into it. Theoretically, the easier fix should just involve parsing all
of the -device commands and looking for a "bootindex=1" field. The Qemu options
code already handles a bulk of this work, so it's just a matter of putting it all
together.Quoted 2 lines
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool
s390_gen_initial_iplb(S390IPLState *ipl){
DeviceState*dev_st;Quoted 7 lines
dev_st=get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default
disk.Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu
options present means we fallback to using zipl values for CCW being tied into the switch statement
is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once
I have something good.
Quoted 7 lines
Quoted 2 lines
This should do the trick (this can also be squished painlessly into 6/13 if desired)
Patch looks fine to me. I can either take it directly like this, or in case you have to respin (depends on the problem that Christian reported with the Ubuntu guest), I'm also fine if you squash it into an earlier patch instead.
FWIW, my problem (a menu happens even without -boot menu=on or loadparm) also happens with other guests (e.g. fedora).
Quoted 80 lines · Christian Borntraeger
Quoted 78 lines · Thomas Huth
Quoted 77 lines · Collin L. Walling
Quoted 76 lines · Collin L. Walling
Quoted 60 lines · Christian Borntraeger
Quoted 51 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 22 lines
Quoted 22 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer
scenario (no libvirt)).If you start QEMU manually without a bootindex, the -boot menu=on is ignored
if no drive has a bootindex.For example:
-drive file=/dev/dasda,if=none,id=d1 -device
virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on
does work-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 15 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently
in the BIOS up to QEMU (find the first defined virtio disk and select it
as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB.
While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]I can look into it. Theoretically, the easier fix should just involve parsing all
of the -device commands and looking for a "bootindex=1" field. The Qemu options
code already handles a bulk of this work, so it's just a matter of putting it all
together.Quoted 2 lines
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool
s390_gen_initial_iplb(S390IPLState *ipl){
DeviceState*dev_st;Quoted 7 lines
dev_st=get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default
disk.Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu
options present means we fallback to using zipl values for CCW being tied into the switch statement
is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once
I have something good.
Quoted 10 lines
Quoted 6 lines
Quoted 2 lines
This should do the trick (this can also be squished painlessly into 6/13 if desired)
Patch looks fine to me. I can either take it directly like this, or in case you have to respin (depends on the problem that Christian reported with the Ubuntu guest), I'm also fine if you squash it into an earlier patch instead.
FWIW, my problem (a menu happens even without -boot menu=on or loadparm) also happens with other guests (e.g. fedora).
Is this on guests using DASD? If so, a boot menu will appear if the appropriate zipl configuration values are set (even if no -boot menu or loadparm). Patch 12/13 goes into more detail.
Is this not wanted behavior?
Signature · 1 line
- Collin L Walling
Quoted 88 lines · Collin L. Walling
Quoted 87 lines · Christian Borntraeger
Quoted 85 lines · Thomas Huth
Quoted 78 lines · Collin L. Walling
Quoted 69 lines · Collin L. Walling
Quoted 60 lines · Christian Borntraeger
Quoted 51 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 22 lines
Quoted 22 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer
scenario (no libvirt)).If you start QEMU manually without a bootindex, the -boot menu=on is ignored
if no drive has a bootindex.For example:
-drive file=/dev/dasda,if=none,id=d1 -device
virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on
does work-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 15 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently
in the BIOS up to QEMU (find the first defined virtio disk and select it
as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB.
While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]I can look into it. Theoretically, the easier fix should just involve parsing all
of the -device commands and looking for a "bootindex=1" field. The Qemu options
code already handles a bulk of this work, so it's just a matter of putting it all
together.Quoted 2 lines
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool
s390_gen_initial_iplb(S390IPLState *ipl){
DeviceState*dev_st;Quoted 8 lines
dev_st=get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the
BIOS will IPL the default
disk.Quoted 8 lines
Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu
options present means we fallback to using zipl values for CCW being tied into the switch statement
is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once
I have something good.Quoted 2 lines
This should do the trick (this can also be squished painlessly into 6/13 if desired)
Patch looks fine to me. I can either take it directly like this, or in case you have to respin (depends on the problem that Christian reported with the Ubuntu guest), I'm also fine if you squash it into an earlier patch instead.
Quoted 9 lines
Quoted 4 lines
FWIW, my problem (a menu happens even without -boot menu=on or loadparm) also happens with other guests (e.g. fedora).
Is this on guests using DASD? If so, a boot menu will appear if the appropriate zipl configuration values are set (even if no -boot menu or loadparm). Patch 12/13 goes into more detail.
Is this not wanted behavior?
Just noticed other email chain -- ignore me.
Signature · 1 line
- Collin L Walling
Quoted 87 lines · Collin L. Walling
Quoted 86 lines · Christian Borntraeger
Quoted 84 lines · Thomas Huth
Quoted 77 lines · Collin L. Walling
Quoted 68 lines · Collin L. Walling
Quoted 60 lines · Christian Borntraeger
Quoted 51 lines · Collin L. Walling
Quoted 3 lines · Viktor Mihajlovski
Quoted 2 lines · Christian Borntraeger
Series
Thanks!!!
Quoted 22 lines
Quoted 22 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer
scenario (no libvirt)).If you start QEMU manually without a bootindex, the -boot menu=on is ignored
if no drive has a bootindex.For example:
-drive file=/dev/dasda,if=none,id=d1 -device
virtio-blk-ccw,drive=d1,bootindex=1 -boot menu=on
does work-drive file=/dev/dasda -boot menu=on
does notinstead it prints:
qemu-system-s390x: boot menu is not supported for this device type.Quoted 15 lines
Quoted 2 lines
and the boots up the default entry.
That should indeed be a separate patch, as it would move logic currently
in the BIOS up to QEMU (find the first defined virtio disk and select it
as boot disk).
In fact it's more complicated than that, because it would have to properly account for -boot order=[acdn] and produce the respective IPLB.
While it makes sense, I wouldn't rush that in but rather change the error message to indicate that -device bootindex is needed to activate the menu, at least for the time being.
[...]I can look into it. Theoretically, the easier fix should just involve parsing all
of the -device commands and looking for a "bootindex=1" field. The Qemu options
code already handles a bulk of this work, so it's just a matter of putting it all
together.Quoted 2 lines
Shall I whip something up and post what I have as a reply to this email chain?
In fact, it should already be there.
static bool
s390_gen_initial_iplb(S390IPLState *ipl){
DeviceState*dev_st;Quoted 7 lines
dev_st=get_boot_device(0);--> if this returns 0 we have no bootindex statement anywhere and the BIOS will IPL the default
disk.Quoted 8 lines
Makes sense. I'm working on making this patch look as clean as possible. The fact that no boot menu
options present means we fallback to using zipl values for CCW being tied into the switch statement
is making things a bit tricky. Just have to think the logic through a bit. Will get back to you once
I have something good.Quoted 2 lines
This should do the trick (this can also be squished painlessly into 6/13 if desired)
Patch looks fine to me. I can either take it directly like this, or in case you have to respin (depends on the problem that Christian reported with the Ubuntu guest), I'm also fine if you squash it into an earlier patch instead.
Quoted 8 lines
Quoted 4 lines
FWIW, my problem (a menu happens even without -boot menu=on or loadparm) also happens with other guests (e.g. fedora).
Is this on guests using DASD? If so, a boot menu will appear if the appropriate zipl configuration values are set (even if no -boot menu or loadparm). Patch 12/13 goes into more detail.
Is this not wanted behavior?
Its on an image file with a scsi bootmap. And for that it is absolutely not wanted to wait forever during boot even if you did not specify to have a boot menu. That will break all existing guest definitions that boot up unattended. Imagine to have lets say 1000 guests with services running. Do you expect the host admin to login to every guest just to start it after the host has rebooted.
Quoted 2 lines · Christian Borntraeger
Series
OK, thanks.
I can pick up the patches from this series and fix the comment in patch 5 and the missing break in patch 13 manually, no need to resend so far.
Quoted 4 lines
menu on scsi and dasd bootmaps tested successfully.
There is one thing that we might want to fix (can be an addon patch since this is a non-customer scenario (no libvirt)).
I agree that this can be an add-on patch.
Thomas
Hmmm, on my ubuntu 16.04 guest, I get the boot menu with no timeout even if I do not specify loadparm or boot menu:
qemu-kvm -drive file=/var/lib/libvirt/qemu/image.zhyp140,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1 LOADPARM=[........]
Using virtio-blk.
Using SCSI scheme.
s390x Enumerated Boot Menu.
3 entries detected. Select from boot index 0 to 2.
Please choose:
This will break setups that want to boot up unattended.
Quoted 170 lines · Collin L. Walling
Due to the introduction of the QemuIplParameter block and taking some time to revist some areas, a few chunks of code have been reworked to better align with those changes. I also think the reworkings improve readability. The same functionality is still there, the code just looks a little different (and hopefully looks better).
--- [v8] ---
The tl;dr:
cleaned up some early patches based on review, threw zipl boot option code into its own patch, refactored
s390_ipl_set_boot_menuto reflect latest changes.The "pls explain":
* cleaned up "s390-ccw: refactor boot map table code"
- removed void ptr casting in a couple of spots
* cleaned up "s390-ccw: set
cp_receivemask..." patch- setting the mask consumes a service interrupt, so there is no need for a followup
sclp_print* fixed uitoa based on feedback from v7
* fixed alignment concerns in QemuIplParameters and renamed flags field
- reasoning: necessary; better readability
- s/boot_menu_flags/qipl_flags so this field can be (potentially) used for other purposes in the future
- when setting the boot menu parms in the bios, we take care to mask out the appropriate
qipl_flagsfor boot menu specific flags- boot menu flags defines are renamed to be prefixed with "
QIPL_FLAG_"- also updated comment above this struct
* cleaned up "s390-ccw: read user input..." patch
- defines for low core external interrupt code addr and clock comparator interrupt code
- take care to make sure buf remains null terminated when passed to
read_prompt* [NEW PATCH] "s390-ccw: use zipl values when no boot menu options are present"
- reasoning: better readability; further explanation of feature
- *nothing new added to this patch series here*
- zipl options flag setting and parsing *moved to* this patch
- this attempts to better explain how these fields are used and how they get parsed
- the commit message gives details on how to set these fields in the zipl configuration file
- the zipl options are only set for CCW type IPL devices (since no other devices actually support it)
* refactored
s390_ipl_set_boot_menu- reasoning: better readability
- the idea is that we should take care to appropriately set the boot menu flags for each IPL device type from the beginning. We should not set certain flags for devices that cannot support certain features (eg SCSI does not support zipl menus, so we should never set the
use_zipl_optsflag for SCSI)- since there are no longer boot menu fields specific to each IPL type, the switch statement is simply used to detect if the IPL device type is capable of a boot menu
- since zipl flags are only set for CCW type IPL devices, I reworked the logic and removed some indentation
* s/menu_check_flags/menu_is_enabled
- reasoning: better readability
- no parameters
- "if menu is enabled" reads better than "if these specific flag bits are set"
* removed
menu.h- reasoning: file not needed; less to maintain
- introduction of qipl and better understanding of zipl options led to this decision. by the end of these changes, this file ended up housing 4 function declarations and no longer seemed necessary
- all menu related function declarations are in
s390-ccw.h- boot menu flags are defined in
iplb.h(which aligns withhw/s390x/ipl.h)--- [Summary] ---
These patches implement a boot menu for ECKD DASD and SCSI guests on s390x. The menu will only appear if the disk has been configured for IPL with the zIPL tool and with the following QEMU command line options:
-boot menu=on[,splash-time=X] and/or -machine loadparm='prompt'
The following must be specified for the device to be IPL'd from:
-device ...,bootindex=1or via the following libvirt domain xml:
<os> <bootmenu enable='yes' timeout='X'/> </os> or <disk> ... <boot order='1' loadparm='PROMPT'/> </disk>Where X is some positive integer representing time in milliseconds.
<boot order='1' ... > must be specified for the device to be IPL'd from
A loadparm other than 'prompt' will disable the menu and just boot the specified entry.
If no boot options are specified, we will attempt to use the values provided by zipl (ECKD DASD only).
Collin L. Walling (13):
s390-ccw: refactor boot map table code
s390-ccw: refactoreckd_block_numto use CHS
s390-ccw: refactor IPL structs
s390-ccw: update libc
s390-ccw: move auxiliary IPL data to separate location
s390-ccw: parse and set boot menu options
s390-ccw: set up interactive boot menu parameters
s390-ccw: read stage2 boot loader data to find menu
s390-ccw: print zipl boot menu
s390-ccw: read user input for boot index via the SCLP console
s390-ccw: setcp_receivemask only when needed and consume pending
service irqs
s390-ccw: use zipl values when no boot menu options are present
s390-ccw: interactive boot menu for scsihw/s390x/ipl.c | 68 ++++++++++++- hw/s390x/ipl.h | 31 +++++- pc-bios/s390-ccw/Makefile | 2 +- pc-bios/s390-ccw/bootmap.c | 184 +++++++++++++++++++++++---------- pc-bios/s390-ccw/bootmap.h | 91 ++++++++++------- pc-bios/s390-ccw/iplb.h | 24 ++++- pc-bios/s390-ccw/libc.c | 88 ++++++++++++++++ pc-bios/s390-ccw/libc.h | 37 ++++++- pc-bios/s390-ccw/main.c | 49 +++++---- pc-bios/s390-ccw/menu.c | 243 ++++++++++++++++++++++++++++++++++++++++++++ pc-bios/s390-ccw/s390-ccw.h | 9 ++ pc-bios/s390-ccw/sclp.c | 39 ++++--- pc-bios/s390-ccw/virtio.c | 2 +- 13 files changed, 741 insertions(+), 126 deletions(-) create mode 100644 pc-bios/s390-ccw/libc.c create mode 100644 pc-bios/s390-ccw/menu.c
Quoted 4 lines · Christian Borntraeger
Hmmm, on my ubuntu 16.04 guest, I get the boot menu with no timeout even if I do not specify loadparm or boot menu:
qemu-kvm -drive file=/var/lib/libvirt/qemu/image.zhyp140,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1
I can reproduce this now, too. FWIW: It only happens with virtio-blk-ccw, not with virtio-scsi-ccw (that's why I did not notice it first). Collin, can you reproduce this, too?
Thomas
Quoted 4 lines · Thomas Huth
Quoted 3 lines · Christian Borntraeger
Hmmm, on my ubuntu 16.04 guest, I get the boot menu with no timeout even if I do not specify loadparm or boot menu:
Quoted 8 lines
Quoted 1 line
qemu-kvm -drive file=/var/lib/libvirt/qemu/image.zhyp140,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1
I can reproduce this now, too. FWIW: It only happens with virtio-blk-ccw, not with virtio-scsi-ccw (that's why I did not notice it first). Collin, can you reproduce this, too?
Thomas
The idea is to support the zipl boot menu on ECKD disks (which have a timeout by themselves)
even without switching the boot menu on. This is to have the same behavior as one would see on LPAR or z/VM.
The issue is that the boot code tries to interpret the program table of SCSI bootmaps as if a boot menu has been specified.
The following patch fixes this. Collin might want to have a say in this matter as well.
The scsi program table was erroneously evaluated as implicit boot menu. This patch adds a per-bootmap-type menu_is_enabled function.
--- pc-bios/s390-ccw/bootmap.c | 4 ++-- pc-bios/s390-ccw/menu.c | 7 ++++++- pc-bios/s390-ccw/s390-ccw.h | 3 ++- 3 files changed, 10 insertions(+), 4 deletions(-)
pc-bios/s390-ccw/bootmap.c +2 -2
@@ -265,7 +265,7 @@ static void run_eckd_boot_script(block_number_t bmt_block_nr,BootMapTable *bmt = (void *)sec;BootMapScript *bms = (void *)sec; -if (menu_is_enabled ()) { +if (menu_is_enabled_for_zipl ()) { loadparm =eckd_get_boot_menu_index (s1b_block_nr); }
@@ -568,7 +568,7 @@ static void ipl_scsi(void)debug_print_int ("program table entries" , program_table_entries);IPL_assert (program_table_entries !=0 ,"Empty Program Table" ); -if (menu_is_enabled ()) { +if (menu_is_enabled_for_enum ()) { loadparm =menu_get_enum_boot_index (program_table_entries); }
pc-bios/s390-ccw/menu.c +6 -1
@@ -237,7 +237,12 @@ void menu_set_parms(uint8_t boot_menu_flag, uint32_t boot_menu_timeout) timeout = boot_menu_timeout; } -bool menu_is_enabled (void ) +bool menu_is_enabled_for_zipl (void ) {return flag; } + +bool menu_is_enabled_for_enum (void ) +{ +return flag & ~QIPL_FLAG_BM_OPTS_ZIPL ; +}
pc-bios/s390-ccw/s390-ccw.h +2 -1
@@ -89,7 +89,8 @@ void zipl_load(void);/* menu.c */ void menu_set_parms (uint8_t boot_menu_flag,uint32_t boot_menu_timeout); -bool menu_is_enabled (void ); +bool menu_is_enabled_for_zipl (void ); +bool menu_is_enabled_for_enum (void );int menu_get_zipl_boot_index (const char *menu_data);int menu_get_enum_boot_index (int entries);
Signature · 6 lines
1.9.1 -- Regards, Viktor Mihajlovski
Quoted 11 lines · Viktor Mihajlovski
Quoted 10 lines · Thomas Huth
Quoted 4 lines · Christian Borntraeger
Hmmm, on my ubuntu 16.04 guest, I get the boot menu with no timeout even if I do not specify loadparm or boot menu:
qemu-kvm -drive file=/var/lib/libvirt/qemu/image.zhyp140,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1
I can reproduce this now, too. FWIW: It only happens with virtio-blk-ccw, not with virtio-scsi-ccw (that's why I did not notice it first). Collin, can you reproduce this, too?
Quoted 6 lines
Quoted 2 lines
Thomas
The idea is to support the zipl boot menu on ECKD disks (which have a timeout by themselves)even without switching the boot menu on. This is to have the same behavior as one would see on LPAR or z/VM.
The issue is that the boot code tries to interpret the program table of SCSI bootmaps as if a boot menu has been specified.
Ok, thanks for the explanation, makes sense now.
Collin, could you please spin a v9 with the required changes included?
Thanks,
Thomas
Quoted 17 lines · Thomas Huth
Quoted 10 lines · Viktor Mihajlovski
Quoted 9 lines · Thomas Huth
Quoted 4 lines · Christian Borntraeger
Hmmm, on my ubuntu 16.04 guest, I get the boot menu with no timeout even if I do not specify loadparm or boot menu:
qemu-kvm -drive file=/var/lib/libvirt/qemu/image.zhyp140,if=none,id=d1 -device virtio-blk-ccw,drive=d1,bootindex=1
I can reproduce this now, too. FWIW: It only happens with virtio-blk-ccw, not with virtio-scsi-ccw (that's why I did not notice it first). Collin, can you reproduce this, too?
Quoted 6 lines
Quoted 2 lines
Thomas
The idea is to support the zipl boot menu on ECKD disks (which have a timeout by themselves)even without switching the boot menu on. This is to have the same behavior as one would see on LPAR or z/VM.
The issue is that the boot code tries to interpret the program table of SCSI bootmaps as if a boot menu has been specified.
Quoted 8 lines
Ok, thanks for the explanation, makes sense now.
Collin, could you please spin a v9 with the required changes included?
Thanks,
Thomas
I see. I overlooked that we could treat a SCSI device as a CCW device. The proposed change makes sense. I'll respin 9 and reply to this chain.
Signature · 1 line
- Collin L Walling