search for: name_to_dev_t

kinit/do_mounts.c:name_to_dev_t() has an inverted logic problem and also doesn't return the correct value for "/dev/nfs". --- klibc-0.194/kinit/do_mounts.c.orig 2005-01-05 01:59:29.250868616 -0700 +++ klibc-0.194/kinit/do_mounts.c 2005-01-05 02:34:30.486432256 -0700 @@ -103,7 +103,7 @@ if ( name[0] == '/'...

...e(resume_file); + fprintf(stderr, "Usage: %s /dev/<resumedevice>\n", progname); + exit(1); } -int resume(const char *resume_file) +int main(int argc, char *argv[], char *envp[]) { - char device_string[64]; - int powerfd = -1; - dev_t resume_device; - int len; - - resume_device = name_to_dev_t(resume_file); - - if (major(resume_device) == 0) { - fprintf(stderr, "Invalid resume device: %s\n", resume_file); - goto failure; - } - - if ((powerfd = open("/sys/power/resume", O_WRONLY)) < 0) - goto fail_r; - - len = snprintf(device_string, sizeof device_string, "%...

...r/kinit/resume/resumelib.c +++ b/usr/kinit/resume/resumelib.c @@ -45,7 +45,9 @@ int resume(const char *resume_file, unsigned long long resume_offset) { dev_t resume_device; int powerfd = -1; + int offsetfd = -1; char device_string[64]; + char offset_string[64]; int len; resume_device = name_to_dev_t(resume_file); @@ -55,13 +57,23 @@ int resume(const char *resume_file, unsigned long long resume_offset) goto failure; } + /* Write the offset, this won't work on less than 4.17 */ + offsetfd = open("/sys/power/resume_offset", O_WRONLY); + if (offsetfd > 0) { + len = snprint...

Hello, There seems to be problem in the cpio unpack in the kernel in initramfs. I have an empty "/sys" directory in my cpio archive, but it never shows up in my extracted initramfs. I've confirmed that extracting the archive again with cpio results in a correct tree, but it doesn't appear on boot when extracted by the kernel. Any ideas? I poked a bit, but I haven't

...r/kinit/resume/resumelib.c @@ -44,8 +44,8 @@ int do_resume(int argc, char *argv[]) int resume(const char *resume_file, unsigned long long resume_offset) { dev_t resume_device; - int powerfd = -1; - char device_string[64]; + int attr_fd = -1; + char attr_value[64]; int len; resume_device = name_to_dev_t(resume_file); @@ -55,30 +55,50 @@ int resume(const char *resume_file, unsigned long long resume_offset) goto failure; } - if ((powerfd = open("/sys/power/resume", O_WRONLY)) < 0) - goto fail_r; + if ((attr_fd = open("/sys/power/resume_offset", O_WRONLY)) < 0) + go...

.../init cmd-arg 1: console=ttyS0,115200n81 cmd-arg 2: root=/dev/nfs cmd-arg 3: ip=::::soekris:eth0 cmd-arg 4: panic=5 cmd-arg 5: cmd-arg 6: BOOT_IMAGE=vmlinuz-2.6.17-rc4-mm1-sk Running ipconfig IP-Config: argc == 2 argv[1]: '-n' IP-Config: no devices to configure kinit: do_mounts kinit: name_to_dev_t(/dev/nfs) = dev(0,255) kinit: root_dev = dev(0,255) Ro[ 25.030001] Kernel panic - not syncing: Attempted to kill init! ot-NFS: no serve[ 25.037537] r defined <0>Rebooting in 5 seconds.. I think this qualifies as a bug; its not what folks expect. A docpatch WITH LARGE LETTERS would rect...

I need to reliably boot a server from a USB device. Since USB device ordering can be unpredictable, I wrote a simple early-userspace "init" program to find the root FS on the correct USB device & partition. I have "labeled" the root ext2 FS on the USB drive, and the program searches for the label by examining the bytes at a specific offset from the beginning of the

This patch series allows user-specified mount commands to be sent in via kernel command line ("kinit_mount=...") or via an embedded /etc/fstab file. The first patch is a cleanup of a patch sent last November by San Mehat (http://web.archiveorange.com/archive/v/EazJNBMORV2U7E0coh5h); the next two are small improvements or bug fixes.

The following patchset implements 3 seperate series of changes. External Building ================= Patches 1 through 8 enable to use of klibc's build system while leaving the src tree pristine (and potentially read only). Specifically: - srctree=<Sources for klibc> - objtree=<Ouput directory for klibc> - KLIBCKERNELSRC=<Kernel sources> - KLIBCKERNELOBJ=<Kernel

Are these the current functional procedures for dropping a dom0 on a jaunty 64 bit server? xen-tools: http://bderzhavets.wordpress.com/2009/08/14/attempt-of-prevu-xen-3-4-1-hypervisor-on-ubuntu-jaunty-server-64-bit/ kernel: http://bderzhavets.wordpress.com/2009/07/07/building-xenified-2-6-30-1-kernel-via-xen-patches-2-6-30-2-tar-bz2/ Regarding the patches used in the above procedure from here:

...; - } -#endif #ifdef CONFIG_BLK_DEV_FD if (MAJOR(ROOT_DEV) == FLOPPY_MAJOR) { /* rd_doload is 2 for a dual initrd/ramload setup */ @@ -352,10 +330,6 @@ { int is_floppy; - mount_devfs(); - - md_run_setup(); - if (saved_root_name[0]) { root_device_name = saved_root_name; ROOT_DEV = name_to_dev_t(root_device_name); @@ -363,6 +337,13 @@ root_device_name += 5; } + if (ROOT_DEV == Root_InitRamFS) + goto done; + + mount_devfs(); + + md_run_setup(); + is_floppy = MAJOR(ROOT_DEV) == FLOPPY_MAJOR; /* This has to be before mounting root, because even readonly mount of reiserfs would...

I have updated the klibc patchset based on feedback received. In particular, the patchset has been reorganized so as not to break git-bisect. Additionally, this updates the patch base to 2.6.17-git12 (d38b69689c349f35502b92e20dafb30c62d49d63) and klibc 1.4.8; the main difference on the klibc side is removal of obsolete code. This is also available as a git tree at: